88 Gametogenesis of the Gall-Fly , Neuroterus lenticularis ( Spathe-gaster baccarum).\#151 ; Part 1 .
By Leonard Doncaster , M.A. , Lecturer on Zoology , Birmingham University .
( Communicated by Prof. W. Bateson , F.R.S. Received September 21 , \#151 ; Read December 9 , 1909 .
) [ Plates 1\#151 ; 3 .
] In previous papers* I have described the maturation of the egg , with some observations on the spermatogenesis , of the Saw-fly , ribesii , in which the eggs may develop either parthenogenetically or after fertilisation .
Some questions remained obscure or doubtful , and it seemed probable that an answer to them might be found in a study of one of the gall-flies ( Cynipidse ) .
The gall-flies are remarkable in having in most species two generations in the year , of which one is bisexual , and the other consists wholly of females .
The flies of the two generations are , further , so different from one another in structure that in most species they were originally described as belonging to different genera .
The galls produced by the two generations are also very distinct .
The species chosen for this work was the very common Neuroterus lenticularis , of which the summer ( bisexual ) generation was originally described as Spathegaster baccarum .
The galls from which the spring ( agamic ) generation emerge are lenticular growths found on the underside of oak leaves in October .
From these galls the flies hatch early in April ; they are exclusively females , and , if provided with buds of oak , readily lay their eggs deep down among the developing leaves and catkins .
The eggs were dissected out , fixed in Petrunkewitsch 's mixture ( alcohol-acetie-sublimate ) at various times after being laid , and cut into sections .
The galls of the summer form are spherical , sappy galls , found on the leaves and catkins in May and June .
From them hatch males and females ( the latter in nature largely preponderating in number ) , and after copulation the females deposit their eggs in the tissue of young leaves , always at the side of a small vein .
The females of this generation differ considerably from the agamic females of the spring brood .
The most conspicuous difference is in the ovipositor , which in the summer females is quite short , but in the agamic females is fully five times as long , and coiled up in the abdomen .
The flies lay readily , and the eggs may be seen with a lens , so that small * 'Q.J.M.S .
, ' vol. 49 , 1906 , p. 561 , and vol. 51 , 1907 , p. 101 .
Gametogenesis of the Gall-Fly , etc. pieces of the leaf containing them are easily cut out and fixed .
Petrunke-witsch 's solution was used , Flemming and others that were tried proving unsatisfactory .
For spermatogenesis and the development of the egg before deposition the larvse and pupae were removed from the galls at the end of May and early in June , opened and fixed immediately in Flemming ( strong ) and Petrunke-witsch .
For this work Flemming 's solution proved the better , but both fixatives gave good results .
The chief stain used in all cases was Heidenhain 's iron-haematoxylin , saf'ranin being used as a control .
Since both males and females arise from the parthenogenetie eggs of the spring generation , it seemed desirable to determine whether both sexes arise from eggs laid by one female , and whether there are any differences in the maturation of the eggs corresponding to the difference in sex .
With these objects in view , in collecting and preserving the eggs , in 1909 I kept those laid by individual females separate ; and I also made some experiments in letting the flies lay on growing oak branches , and rearing the galls .
In April I put single females in muslin bags or " sleeves " on the branches of an oak tree ; removed the sleeves and marked the branches when the flies had died , and searched the leaves for galls at the beginning of June .
In 1908 these experiments were wholly unsuccessful , perhaps because the tree chosen had not suitable buds for the flies to lay in , for in 1909 I found that they laid much more readily in buds from some trees than from others .
The more advanced buds were preferred .
In 1909 I sleeved out 40 flies on several different trees , including some quite young oaks from 1 to 4 feet in height .
On June 1 , I found galls on seven branches of two large oaks , but none on any of the small ones .
The numbers of galls in the sleeved branches were respectively 6 , 7 , 7 , 2 , 4 , 3 , 3 , and from these I reared respectively 5 ?
's , 7 \lt ; $ 's , 6 's , 2 ?
's , 4 $ 's , 0 , 0 .
In the galls which did not hatch the larvae either died or were parasitised .
I searched the trees carefully for galls on branches which had not been sleeved , and found none , so it may be concluded with confidence that these galls were produced by the flies sleeved out .
Since in each case only one sex emerged from the galls produced by one fly , although the numbers are small I think it is justifiable to conclude that every fly of the agamic generation produces eggs of only one sex , i.e. gives origin to either males or females only , not to both .
Cases of this kind are known in the Hemiptera ( Aphididse ) and in Rotifers , but not hitherto , as far as I am aware , in the Hymenoptera .
Mr. L. Doncaster .
[ Sept. 21 , The Somatic Mitoses .
In attempting to count the chromosomes in the mitoses in the egg , it was found difficult to determine the exact number , and with a view to checking this a number of counts were made of division-figures in the body tissues ( developing wings , hypodermis ) of male and female pupae .
Some of these mitotic figures are represented in Plate 1 , figs. 1 and 2 .
In all cases the chromosomes are nearly equal in size , and the number is certainly approximately 20 in somatic mitoses .
In mitoses in male pupse or larvae just before pupation , 50 counts gave an average of 19'2 chromosomes , 17 showing 20 , and the remainder 19 or 18 .
In female pupae of the summer generation , 40 counts gave an average of 19'4 , 25 showing 20 , and the rest 19 or 18 .
In the pupae ( female ) of the spring generation , division-figures are scarce , as development proceeds much less quickly , but it is clear that the number is sensibly the same as in the summer brood ( Plate 1 , fig. 3 ) .
The fact that in 50 counts of male mitoses only 17 showed 20 chromosomes , while 25 out of 40 had that number visible in females , might suggest the possibility that the male actually has one chromosome less than the female .
In cases such as this , where the chromosomes are small and crowded , it is quite possible to count one elongated or bent chromosome as two , but some of the mitoses with 20 are so clear as to leave no doubt that this is the true number .
The difference probably depends on the fact that the female pupa from which most of the counts were made had exceptionally line mitoses , and was very well preserved , so as to give a smaller proportion of irregular figures than in the male pupae used .
It is possible that in the male one chromosome is smaller than the rest , and sometimes hidden , but I have not been able to convince myself of this .
Spermatogenesis .
The larvae are found in the galls in the latter half of May ; they develop very rapidly , for at the middle of the month they are exceedingly small , and by the end most of the galls contain pupae .
In very young pupae , with no colour except in the eyes , the testes contain spermatocytes and the various maturation phases ; a week later all the pupae are grey or black , and the " meiotic phase " is over , only spermatids and nearly ripe spermatozoa being found .
The full-grown larvae also have spermatocytes in the testes , and some follicles show prophases of the maturation ; for the spermatogonial stages , half-grown larvae , about four or five days younger , are required .
In very young larvae I have not been able to distinguish the gonads .
The larvae and pupae were opened with needles and fixed entire in 1909 .
] Gametogenesisof the , etc. Flemming 's strong fluid , or sometimes in Petrunkewitseh 's fluid ; owing to their small size it was found that when the animals were well opened , fixation was quite successful after half an hour in strong Flemming , and if they remained too long the cells became over-fixed .
The sections were stained in Heidenhain 's iron-hsematoxylin and sometimes in safranin as a control .
In larva ?
, shortly before pupation , the testes contain only spermatocytes , and no division-figures are found .
In younger larva } it is not quite easy to distinguish the sexes , but the ovaries are larger , the cells are beginning to arrange themselves in strands to form the egg-tubes , and some cells are larger than the rest , with very large nuclei , and presumably are developing into eggs .
The testes are smaller and all the cells appear alike .
They have vesicular nuclei with a large nucleolus , and are much like the embryonic cells of the body-tissues .
I have found only one larva* which shows the spermatogonial divisions , and these are remarkable from the large size of the chromosomes and spindle , and from the fact that there are clearly 10 chromosomes at each end in the anaphase ( Plate 1 , fig. 4 , b , c ) .
The chromosomes are elongated and rod-like , and when the spindle is cut across , the number 10 can be counted with great confidence at each end .
In the same larva mitoses in the body-cells show clearly the diploid number ( about 20 ) ; in this particular larva I have found none in which I can count the chromosomes with perfect accuracy , but in many mitoses at least 18 can be seen without any doubt , both in metaphase and anaphase .
It .appears , then , that while the germ-cells contain only 10 chromosomes at this stage , the body-cells contain 20 .
In female pupse of the same age , as will be described below , 20 chromosomes appear both in oogonial and somatic divisions .
In larvte , shortly before pupation , the testes are like those of young pupa } , ?
but contain only primary spermatocytes ; in some follicles these are in the * Since the paper was sent for publication , I have found a second young larva showing 10 chromosomes in the spermatogonial mitoses both in metaphase and anaphase .
I have also discovered that in the developing central nervous system of male larva ; ( several different specimens ) many , but not certainly all , of the mitotic figures show the haploid number of chromosomes ( 10 ) .
In female larvae the mitoses of the nervous system contain the diploid number .
I have not found any perfectly clear equatorial plate in the male nervous tissue showing the diploid number , but some figures suggest it , and it is possible that the supporting cells , like those of the hypodermis , have the full number , and the nerve-cells the reduced .
It should also be mentioned that in both sexes there occur in places , in or just below the hypodermis , giant nuclei some 15 ju .
or 20 in diameter , and in one female larva I have found a division figure of one of these nuclei in anaphase with at least 50 chromosomes at each pole .
A fuller description , with figures , will be given in the second part of this paper.\#151 ; Novembe , 1909 .
Mr. L. Doncaster .
[ Sept. 21 , resting condition , but in others chromatin is beginning to appear , and where it is aggregated into distinct masses the number of these approaches 20 , but they are too irregular for an accurate count to be possible .
The testes in the young pupse are divided into follicles , in the walls of which are very large nuclei with many chromatin granules .
In each follicle the cells are nearly , but not quite , exactly at the same stage of development .
This fact makes it very difficult to determine with certainty the exact course of events , for only here and there does one find a follicle in which slightly different stages are represented , and so get a clue as to the course of development .
In pupse which are not too far advanced , various stages occur in different follicles of the same testis from primary spermatocytes before the maturation divisions have begun , up to spermatids which are beginning their metamorphosis into spermatozoa .
By comparing a number of testes of this age with one another , I have found a continuous series of stages , and think there is little doubt that in the account which follows they are placed in the correct order .
The difficulty is increased by the fact that stages which appear closely similar occur at considerable intervals , and great care is needed in distinguishing one from the other .
The primary spermatocytes are rounded cells lying free in the cavity of the follicle ; each has a large vesicular nucleus , containing a very faintly-staining reticulum , and often a nucleolus ( Plate 1 , fig. 5 a , At the approach of the " meiotic " divisions chromatin appears as small dots close-to the nuclear membrane ; these dots coalesce into irregular masses just inside the membrane , and these then form themselves into 10 chromosomes , , which in some cases for a short time have an elongated band-like form ( Plate 1 , fig. 6 ) .
The cell is meanwhile becoming elongated and pear-shaped , , the nucleus occupying the wider end .
At the apex of the narrow end a minute black dot , the centrosome ( centriole ) , may often be seen , and at a slightly later stage a second centrosome is seen at the broad end of the cell , close to the nucleus ( fig. 7 a ) .
I have not observed their origin .
In the primary spermatocyte , when the chromatin is beginning to appear , a slightly larger black dot appears outside the nucleus , but as this may still be seen , in addition to the centrosomes , at the stage described , it cannot give origin to the centrosomes .
The chromosomes now become shorter , and place themselves in a radial arrangement , as in an equatorial plate ; the nuclear membrane persists , but the nucleus is drawn out towards the narrow apex of the cell , so that it also becomes somewhat pear-shaped ( Plate 1 , fig. 7 c ) .
At this stage , in some-cells the narrow apex of the cell is elongated into a fine process , with th\amp ; centrosome ( centriole ) at its tip ( fig. 7 c ) , like that figured by Mark and Gametogenesis of the Gall-Fly , etc. 1909 .
] Copeland in the corresponding stage in the Bee.* Fine threads run down from this to the nucleus , but it is difficult to determine whether they penetrate inside the membrane or pass outside it , for at the narrow end the nuclear membrane becomes indistinct and confused with these fibres , while remaining clearly defined at the opposite wider pole .
The stage now described resembles the metaphase of a true division , such as occurs later , but is distinguished from it by the persistent nuclear membrane , and the position of the chromosomes at one end of the nucleus , near the broad end of the cell .
In this case no nuclear division takes place , the nucleus becomes oval in shape , and the chromosomes generally contract to form a compact mass lying across its centre ( figs. 7 8 ) .
In some cells at least this chromatin mass seems to divide , half passing to each end of the oval nucleus , and as this returns to its resting condition it is common to find a cap of chromatic material at each pole .
The chromatin may finally disperse , taking the form of granules , and thus giving rise to a condition not differing greatly in appearance from the primary spermatocyte in which the chromatin has begun to appear .
Possibly the division of the chromatin inside the nucleus , which occasionally seems to occur , is the persistent remnant of a true nuclear division , or it may perhaps be compared with the " intranuclear karyokinesis " described by Kostaneckif in partlienogenetio eggs of Mactra .
While the nucleus is returning to its resting condition , the apex of the pointed end of the cell , with the centrosome , becomes constricted off , apparently much in the same way as in the Bee .
The amount of cytoplasm removed with the centrosome is very small , but it may sometimes be seen attached to the body of the cell by a narrow bridge containing spindle-fibres ( fig. 7 b ) .
In some cases the separation seems to be complete , and in follicles containing spermatocytes with a reconstituted nucleus small loose pieces of cytoplasm , sometimes with the centrosome visible at one end , are found scattered among the cells .
In other cases these fragments seem toj remain attached to the cell , but soon degenerate ; in the prophase of the ; succeeding division they are perhaps represented by small knob-like excrescences sometimes found at the edge of the cell ( figs. 8 , 9 , 10 ) .
The cells with reconstituted nuclei may now be regarded as secondary spermatocytes , although no true nuclear division has occurred .
In their resting condition , the duration of which it is not easy to determine , they do not differ greatly from the primary spermatocytes when the chromatin is beginning to appear ; but they are generally more elongated , * ' Proc. Amer .
Acad. Arts and Sciences , ' vol. 42 , No. 5 , fig. 8 . .
t ' Arch. Mikr .
Anat .
, ' vol. 64 , 1904 , p. 1 .
VOL. LXXXII.\#151 ; B. T Mr. L. Doncaster .
[ Sept. 21 , and very frequently the chromatin is closely packed round the nuclear membrane , especially at one pole , with fine strands radiating from this towards the more empty pole .
At this stage they are much like the spermatids , in which the chromatin has a similar arrangement figs. 8 c and 17 ) , but the nucleus is much larger , having the same size as that of the primary spermatocytes .
The chromatin next becomes grouped in the form of large elongate granules or small bands scattered under the nuclear membrane ; their number is not easy to determine , but it approaches 20 ( fig. 9 ) .
They then become combined , but whether by end-to-end pairing I cannot determine , into 10 very definite long bands having a more or less regular meridional arrangement under the membrane ( fig. 10 a , b , c ) .
In one testis , in which a single follicle is much retarded , all the others containing well advanced spermatids , there are about 20 of these bands in each nucleus , and as they concentrate themselves to form chromosomes they appear to pair side by side to form 10 split chromosomes ( Plate 1 , figs. 18 , b ) .
It is possible that this lagging follicle is abnormal , for in all other cases where the bands can be counted their number is clearly about 10 .
The bands then become shorter , and arrange themselves across the nucleus in a fairly regular equatorial plate , in which 10 chromosomes radiate from a centre ( figs. 11,12 ) .
The nucleus has now assumed the form of a wide spindle , stretching almost from end to end of the cell , but the membrane , though faint , appears still to persist .
At each end of the nucleus a minute centrosome may be seen , and apparently within the nucleus is a system of spindle fibres extending from the centrosomes to the chromosomes , which lie across the centre .
Commonly also at this stage a small deeply-stained dot may be seen outside the nuclear spindle , generally nearer one end .
It is shown in figs. 13 , 14 , 15 ; where , as in figs. 13 b , 14 a , it appears within the spindle , it is in reality at a different level , and is always quite near the edge of the cell .
The metaphase condition appears to last for some little time , for it is frequently found ( fig. 13 In the early anaphase the nuclear membrane has quite disappeared , and the chromosomes appear to split longitudinally , for in the equatorial plate they are arranged with their length across the spindle .
The cell elongates somewhat , and the spindle extends completely from end to end .
In early anaphase the chromosomes have the form of short rods converging towards the centrosomes ( fig. 14 a,6 ) ; as they move apart they shorten , and as they aggregate round the centrosome they become still more concentrated .
The cell meanwhile becomes constricted , a sheaf of spindle fibres extending from one daughter nucleus to the other , connecting the two halves ( fig. 15 The two groups of chromosomes round themselves off into nuclei , having about 1909 .
] Gametogenesisof the , etc. half the size of the spermatocyte nuclei ; the cells divide completely , and become a pair of spermatids ( figs. 16,17 ) .
In late anaphases the small stained body \#166 ; described above outside the nucleus may usually be seen included in one of the daughter cells ; it is found outside the nucleus or spindle during the whole process , so cannot be regarded as any kind of accessory chromosome ( figs. 12 , 13,14 , 15 ) .
It will be seen that the spermatogenesis in presents the same remarkable features as have been described by Meves* and Mark and Copelandf in the Bee and Wasp .
In each we find the first spermatocyte division suppressed , being represented by the extrusion of a centrosome with a small quantity of cytoplasm .
Neuroterus resembles the Wasp in forming two nearly similar spermatids from each spermatocyte , but differs from both Bee and Wasp in having a definite resting stage between the abortive and true divisions .
The presence of the small stained body , persisting from the primary spermatocyte and passing into half the spermatids , has not been recorded in either Bee or Wasp .
Development of the Egg in the Ovary .
In rather young larvae of the summer generation the ovaries consist of masses of cells not yet divided into egg-tubes , but they are beginning to arrange themselves into strands from which the egg-tubes will develop .
No clear distinction is yet visible between the cells which will become eggs and those which will form the accessory cells between the eggs .
The nuclei are mostly large , reticular , with one or two conspicuous nucleoli ; others are stil larger , and have a mass of chromatic material aggregated in the form of dots and fine threads at one side of the nucleus .
Nuclei like this may also be found in the eggs in the pupal egg-tubes .
Mitotic figures are frequent ; they resemble those found in the body-tissues , and clearly have the diploid number of chromosomes ( Plate 1 , fig. 19 a , b ) .
In young pupae of the summer generation , the stages in the development of the ovarian egg can be followed easily , for the upper ends of the egg-tubes contain very early stages , and at the lower end well developed eggs are found .
In the upper ends of the tubes all the cells are alike , but lower down differentiation takes place , and some cells enlarge greatly , become filled with yolk , and begin to develop into eggs surrounded by follicle-cells .
Alternating with the developing eggs are masses of cells with large nuclei , each usually containing a double nucleolus ; these cells are all alike , and do not differ much * ' Arch. Mikr .
Anat .
, ' vol. 70 , 1907 , p. 414 ; vol. 71 , 1908 , p. 571 .
t 'Proc .
Amer .
Acad. Arts and Sciences , ' vol. 42 , 1906 , p. 103 ; vol. 43 , 1907 , p. 71 .
Mr. L. Doncaster .
[ Sept. 21 , rom the primitive ova at the top of the egg-tube .
The tube is thus monili-form , and the swellings contain alternately a developing egg and a mass of cells like primitive ova ( fig. 20 ) .
As the eggs develop they enlarge greatly by the deposition of yolk , and the nucleus appears as a relatively small vesicle , lying at one side .
When ripe , the egg is prolonged at the front end into a hollow stalk with a small vesicle at the free end ; the stalk is said to be held by the ovipositor as the egg is thrust into the hole bored in the leaf when the egg is being laid .
The stalk thus partly blocks the opening , and in the spring eggs , which are generally somewhat compressed by the bud-scales , some of the yolk of the egg is frequently forced into the cavity of the stalk .
After one of the primitive ova has definitely begun to develop into an egg , as shown by the deposition of yolk , no division of the nucleus appears to take place until it is laid .
In the follicle-cells which surround the eggs mitoses are not uncommon , and show the same number of chromosomes as in other somatic nuclei , viz. , about 20 .
In the females of the spring ( agamic ) generation , I have not observed the oogonia .
In the pupae the development does not differ materially from that of the summer generation .
The egg-tubes are very similar , and the eggs when laid differ chiefly in being slightly larger .
Since the pupae develop very slowly , mitoses are more difficult to find , but the chromosome groups from an egg-follicle , represented in fig. 3 , show that the somatic number is similar to that in the summer pupae .
Fig. 21 c , shows three figures of an egg-nucleus in which about 20 chromosomes appear to be visible , and b being drawn from the same section at different levels , and c from the next section of the series .
Maturation and Fertilisation of the Fgg.\#151 ; Summer Generation .
The eggs laid by the flies of the summer generation in the early part of June are sunk in the tissue of the underside of young oak leaves , and the stalk of the egg commonly projects somewhat from the hole made by the ovipositor .
These eggs are fertilised , and the large spermatazoon may almost always be found near the edge in sections of eggs preserved within about an hour and a half from the time of laying .
The spermatozoon , after entering the egg , and while being converted into the male pronucleus , generally lies at the opposite side of the egg , or at least some distance removed , from the egg-nucleus , which during the same period is undergoing its maturation processes .
During the first hour and a half after the egg is laid the nucleus is-found near the edge ; its position is somewhat variable , but it is commonly about midway between the ends of the egg ( Plate 2 , fig. 22 ) .
1909 .
] Gametogenesis of the , In eggs preserved very shortly after being laid , the nucleus appears as a small darkly-stained body , either flattened against the edge of the egg or somewhat spindle shaped , with the longer axis of the spindle perpendicular to the edge .
There is little doubt that the very narrow flattened nucleus is the earlier stage , but as both are found in eggs within half an hour of being laid , this is not perfectly certain .
The nucleus now begins to enlarge , and instead of staining deeply and almost evenly throughout , chromatin bodies connected more or less conspicuously by a network appear within it ( fig. 23 ) .
The succeeding stages are somewhat obscure , and are not unlike the maturation processes described by Henking* in the Gall-fly The nuclear membrane becomes faint and disappears , and the chromatin bodies segregate themselves to some extent into an inner and outer group , but whether by division of the individual chromosomes , or by a separation of chromosomes previously distinct , I cannot be sure ( Plate 2 , figs. 22 , 24 , 27 , 30 ) .
The chromosomes of the inner group then separate themselves from those of the outer , and in so doing take the form of rods lying side by side or with their inner ends converging somewhat as in the anaphase of a typical mitosis ( Plate 2 , fig. 30 ) .
The outer part of the nucleus meanwhile has never become clearly separated into chromosomes , but appears as a group of chromatin bodies connected together by strands as in a reticular nucleus .
Although Henking does not figure a stage exactly of this kind in Bhodites , yet he describes something similar .
On pp. 149 and 150 he writes : " Die untere Tochterplatte lasst die Neunzalil unschwer erkennen , wahrend in der ausseren Tochterplatte wiederum eine theilweise Verklebung eingetreten ist .
" In a nuclear division of this kind , in which there is never an equatorial plate , and in which the chromosomes of one half never become clearly separated , it is not easy to determine the chromosome number with certainty .
But from a comparison of a large number of sections , especially when cut tangentially to the egg , it is fairly clear that the number of chromosomes in the inner group is about ten .
In some sections not so many are visible , but that the number is really approximately ten there can be no doubt , i.e. the same number as in the spermatocyte divisions already described .
In Bhodites Henking found nine .
The division just described does not lead to the production of two resting nuclei , but is succeeded immediately by further changes , which here also are obscure and hard to follow .
Henking says that in Bhodites the chromosomes at each end of the first division-figure again divide , and thus give rise to four groups , of which the innermost forms a definite nucleus which sinks in as the female pronucleus , while the other three remain near * ' Zeit .
Wiss .
Zoo .
, ' vol. 54 , 1892 , p. 147 .
Mr. L. Doncaster .
[ Sept. 21 .
the egg margin as polar bodies .
Such a process would not be very different from what occurs in the saw-flies and other Hymenoptera , but Heinking confesses that he has never actually seen this second polar mitosis .
" Die Theilung muss ausserordentlich rasch verlaufen ; denn . . . .
habe ich hier immer gefunden , dass die Theilung der Chromosomen bereits vollendet war .
" ' Like Henking , I have had difficulty in observing this second polar division , although stages with three groups of chromosomes and a developing pronucleus are abundant .
It is clear that the outer and inner groups left from the first polar division do not divide simultaneously* and although I have many sections of these stages I have no clear figure of the division of the inner group , i.e. of the separation of the female pronucleus from the chromosomes of the second polar body .
For some time I doubted whether any such division occurred , but the presence of three groups of polar chromosomes is difficult to explain without it , and I have obtained a few sections which suggest that such a division is taking place .
It clearly occurs before the division of the outer chromosome group , for while the latter is still a confused mass near the edge of the egg , the inner group often appears very much drawn out ( Plate 2 , figs. 26 , 30 ) as if undergoing division , and in other sections ( fig. 31 ) the division is seen Completed .
This stage seems to follow immediately on the first division , so that the division of the inner group of chromosomes is part of the same process as the original separation into inner and outer groups .
In Plate 2 , fig. 29 , a different phase is represented , in which the inner group looks as if it were forming a compact mitotic figure , but this appearance is not usual .
After the chromosomes which will give rise to the female pronucleus have sunk in to some extent , the outer group of Chromosomes , lying near the edge of the egg , undergoes an irregular division ( Plate 2 , fig. 32 ) , so giving rise to three groups altogether of polar chromosomes , but these are SO confused and irregular that the number in each group is never ascertainable with certainty , and it appears as if some fusion often takes place between them ( fig. 33 ) .
There is One possibility which should be mentioned here , which is not entirely inconsistent with any of my sections .
It is that the three groups of polar chromosomes are all derived by division of the outer group left by the first maturation division .
I have no section which proves with certainty that the inner rod-like chromosomes described above are not converted direct into the female pronucleus , and that the three groups of polar chromosomes are not produced by a separation of the original outer group into two , followed by a second division of the outermost , and so [ 909 .
] Gametogenesisof the , etc. yielding three groups .
If this is the case , the innermost of the three groups must be formed by a sorting out of certain chromosomes from the original outer group , followed by a division of the remainder .
It has been seen that there is no evidence for more than one maturation division in the spermatogenesis , but the existence of only one such division in the egg does .
not seem consistent with the presence of the diploid number of chromosomes in the oogonia .
The further fate of the polar chromosomes seems to vary somewhat in different eggs .
In some , at the time of the conjugation of the male and female pronuclei , they appear as a single large group surrounded by a field clear of yolk granules , and with the individual chromosomes long and thread-like .
Very frequently there are two groups , as if the two inner had amalgamated , and occasionally one finds three or four chromatin masses , in these cases usually with the chromatin closely balled together .
In rather later stages , during the segmentation ( about five to eight hours ) , the polar chromosomes seem to shrivel into small irregular masses and then to disappear completely .
It may perhaps be suggested that the apparent abnormality of the whole of the maturation processes of the egg is due to defective methods of preservation , but I think it unlikely for the following reasons .
In the first place , at a rather later stage the conjugation of the pronuclei and segmentation divisions are well preserved ; and secondly , Henking found very similar abnormalities in Rhodites in eggs preserved by hot water and by Flemming .
It seems unlikely that three so different methods of preservation should all give the same results if the phenomena were not genuine .
The division-figures in the cells of the leaf-tissue are also well fixed .
While the maturation of the egg-nucleus is in progress the spermatozoon is being converted into the male pronucleus .
In eggs preserved within half an hour after being laid it commonly appears as a long narrow rod , straight or slightly curved , and often extending through two or even three sections .
It contracts to a small , oval , compact nucleus , which stains very deeply , and then gradually swells , while chromatin bodies become visible within it .
At this stage the chromatin sometimes suggests a coiled thread , but the staining is too dense to see whether this is so with certainty ( Plate 2 , fig. 32 ) .
As it increases in size , the male pronucleus sinks into the egg to meet the female pronucleus , which has now reached a similar condition , and when they meet in the centre of the egg they are both very large and vesicular ( Plate 3 , fig. 34 ) .
The chromatin masses become definite chromosomes , the nuclear membranes disappear , and the two groups of chromosomes mingle and begin to form Mr. 'Ii , Doncaster .
[ Sept. 21 , the first segmentation mitosis .
Minute centrosomes are visible at the poles of the spindle , but I have not observed their origin .
Sometimes at least the chromosomes in the equatorial plate of this first division are seen to be in two groups side by side , so that the complete mingling cannot take place till the first segmentation nuclei are reconstituted ( fig. 35 ) .
In the later segmentation mitoses the chromosomes are elongated , - and appear to be about 18 to 20 in number ; and in the nuclei just before mitosis a similar number of chromatin bands may be counted ( figs. 36 to 38 ) .
Maturation and Segmentation of the Egg of the Spring Generation .
My observations on the eggs of the spring generation are not yet complete , and cannot be finished until fresh material has been obtained .
I give here a preliminary account of the results , some of which cannot be regarded as established with certainty , and leave the full description until the second part of this paper is published .
The eggs of the spring generation are parthenogenetic ; they are laid in the developing buds of the oak in April , and as has been mentioned above , those laid by some females develop into males , those laid by others to females .
When the maturation divisions have been found they resemble the early stages of the maturation in the summer eggs , but I have never seen the division of the inner chromosome group , and am inclined to believe that it may sink in and become the egg-nucleus without further division , so that only one true maturation division takes place ( Plate 3 , figs. 39 to 43 ) .
In other eggs I have been unable to find any trace of a maturation division , and in the later stages , during segmentation , no polar chromosomes are to be found in such eggs , although they are always to be found in the summer generation at the same stage .
It seems probable , therefore , that thdre are two kinds of eggs , of which one undergoes a maturation division and the other does not , and this is confirmed by the study of the segmentation mitoses .
In the eggs laid by the majority of spring females , 20 chromosomes are found in the segmentation mitoses ( Plate 3 , figs. 44 ; 45,46 a , hi c ) , but in the eggs laid by one female the division-figures all contain 10 or about 10 chromosomes ( figs. 47 , 48 a , e ) , and in some of these eggs a double group of polar chromosomes is clearly recognisable at the edge of the egg in the place where the maturation mitoses have been found at an earlier stage .
Most of my material was collected before I discovered that some individuals lay eggs which yield males , others which yield females , so that I have only one series of eggs laid by one female which show the reduced number in the segmentation divisions .
But as far as they 1909 .
] Gametogenesis of the Gall-Fly , etc. go , the results indicate that one kind of parthenogenetic female lays eggs which undergo no maturation , or at least no reduction , while the other kind lays eggs which undergo a maturation division , and in which the segmentation mitoses show the reduced number ( 10 ) of chromosomes .
It has already been shown that in the male of the summer generation the spermatogonial divisions have 10 chromosomes , although the body-cells have 20 , so that it is probable that the eggs which have a maturation division give rise to males , and it is possible that the 10 chromosomes seen in their segmentation mitoses are bivalent .
In the body-cells these would then split into their univalent components , giving the 20 chromosomes observed , but in the developing germ-cells the haploid number of bivalents would be retained until the single spermatocyte division .
In the summer females , on the other hand , the oogonia show 20 chromosomes , and these are probably produced by females which lay eggs that undergo no reduction .
The complete proof of these facts must be left until a fresh supply of material can be obtained in the spring .
Summary and Discussion .
The chief results described in the foregoing pages are the following:\#151 ; 1 .
The G-all-fly , Neuroterus lenticularis , has two generations in the year , the flies appearing in April and June respectively .
The spring generation consists exclusively of females , which differ considerably from the females of the summer brood .
Their parthenogenetic eggs are laid in oak buds , and all the eggs laid by any one female develop into individuals of the same sex in June , i.e. some of the spring females are male-producing , others are female-producing .
The summer generation thus consists of males and females ; their eggs are fertilised , and are laid in the tissue of young oak leaves , and give rise to galls very different from those produced in the spring .
The flies from these galls hatch in April and thus complete the cycle .
2 .
Mitoses in the body-tissues of youug pupae show about 20 chromosomes , both in the spring parthenogenetic females , and the males and females of the summer brood .
3 .
In the spermatogonia of young male larvae , mitoses show 10 chromosomes .
In the primary spermatocytes of young pupae 10 chromosomes appear .
An imperfect mitotic figure is developed , but the nuclear membrane does not disappear ; and after the metaphase is reached the nucleus returns to a " resting " condition .
During this process the cell develops an elongation at one end , at the tip of which is one of the centrosomes ( or centrioles ) ; as the nucleus re-form3 , the centrosome and a small piece of cytoplasm are separated ofi , as happens in the Bee and Wasp .
The process is thus much Mr. L. Doncaster .
[ Sept. 21 , like that found in the Bee , except that the nucleus returns to a " resting * condition .
; 4 .
The spermatocytes now develop about 20 chromatin masses which form themselves into 10 band-like chromosomes .
These shorten , form an equatorial plate across the cell , a typical spindle is produced ; and the ehromosomes divide so that 10 travel into each daughter nucleus .
Two spermatids are produced , which are similar except that in some cases at least One of them receives a small extranuclear body of unknown nature , which is absent from the other .
Both spermatids develop into spermatozoa .
5 .
In primitive ova in the young female larvae of the summer generation , mitoses like those in the body-cells are found , with apparently 20 chromosomes .
After the deposition of yolk has begun , no further nuclear divisions Occur in the egg .
6 .
The maturation divisions of the summer eggs are difficult to follow , but apparently two divisions occur , giving rise to four groups of chromosomes , of which the three outer represent the three polar nuclei , while the innermost group sinks in to form the female pronucleus .
This group probably consists of 10 chromosomes , but they are so crowded that the number commonly appears rather less in sections .
7 .
The male pronucleus and the female pronucleus meet and form the first segmentation spindle , in which , as in the later segmentation divisions , about 20 chromosomes appear .
The polar ehromosomes disintegrate and disappear .
8 .
The primitive ova of the spring generation have not been observed .
In an egg from the egg-tube of a young pupa 20 chromosomes were seen in the nucleus .
9 .
The maturation of the spring egg has not yet been sufficiently studied , but it appears that some eggs undergo at least one maturation division , others probably none .
In eggs in which maturation has occurred segmentation mitoses show 10 chromosomes ; all the eggs laid by one individual female in which the chromosomes could be counted were of this type , and it is suggested that these develop into males .
In the eggs laid by other females , however , 20 chromosomes appear in the segmentation divisions ; in these , polar chromosomes appear to be absent , and it is probable that there has been no maturation division , and that these eggs would develop into females ; Most of the above facts have been shortly discussed in the sections where they are described , and only their more important general bearings remain to be considered .
Of these , perhaps the chief is the relation of the facts observed* in Neiirdierusto what is already known in other Hymenoptera , 1909 .
] Gametogenesis of the Gall-Fly , etc. especially the Bee and other Aculeata .
The life-history in the gall-flies is more complicated , involving an alternation of two different generations , but in some respects this simplifies our understanding of the facts .
In the Bee and its relatives , fertilisation of the egg is facultative , and to all appearance , at least , the maturation of the egg is similar , whether fertilised or not .
In the gall-flies the parthenogenetic eggs belong to a different generation , and since they give rise to both sexes , sex-determination in this generation , at least , does not depend on fertilisation .
In the summer brood all eggs are fertilised and all give rise to females , so that here the gall-flies resemble the Bee .
As in the Bee also , in the spermatogenesis there is only one maturation division of the nucleus , the number of chromosomes in the spermatogonial mitoses being the same as that entering the spermatids ; in each case the first spermatocyte division is abortive , only a small portion of the cell being extruded .
The Gall-fly differs from the Bee , but resembles the Wasp and Ant in producing two complete spermatids from each spermatocyte , while in the Bee one of the daughter nuclei degenerates without becoming a spermatozoon .
It is tempting to speculate on the reason for this peculiarity in the Hive-bee , and a suggestion with regard to it will be made below after the parthenogenetic generation has been considered .
In the spring generation of Neuroterus , the eggs of which develop parthenogenetically , the evidence is unfortunately not yet complete .
It has been seen that the eggs of some individuals give rise to females , those of others to males .
The microscopical evidence , so far as it goes , suggests that in the eggs laid by some females there is possibly no maturation division , and in any case these eggs contain the diploid number of chromosomes ; in eggs laid by other females a maturation division occurs , and the segmentation mitoses show the halved ( haploid ) number .
It is suggested that the former type of egg develops into a female and the latter into a male , but since there is possibly only one maturation division in these , and since the body-cells of male pupae contain the diploid number of chromosomes , it is possible that the chromosomes of the segmentation mitoses are bivalent .
In the Bee also the body-cells of the male contain more chromosomes than the spermatogonia , in this case apparently four times as many ( about 64 instead of 16).* If it is the fact that parthenogenetic eggs of Neuroterus which have the diploid chromosome number in their segmentation mitoses develop into females , and those which undergo reduction become males , the sex-determination may be imagined as follows :\#151 ; * Meves , 'Arch .
Mikr .
Anat .
, ' vol. 70 , 1907 , p. 420 .
Mr. L. Doncaster .
[ Sept. 21 .
The egg which is to undergo maturation may be regarded as containing both male and female determinants ; the female determinant is removed in the polar mitosis and the egg remains male .
The eggs which will undergo no reduction may be considered as containing only a female determinant , and this causes the egg to develop into a female .
If , as suggested , the ovarian eggs of some parthenogenetic females contain both male and female determinants , while those of others contain a female determinant but not a male , the difference between the two kinds of eggs must be caused by a difference of constitution in the females which lay them .
Both kinds of female are produced from fertilised eggs , and the difference in constitution between them may be due to the existence of two kinds of spermatozoa in the males of the previous generation .
The existence of spermatozoa of two kinds is indicated by a number of facts which will be further discussed below , but here it will suffice to refer to the evidence in the Hymenoptera and other parthenogenetic insects .
In Neuroterus itself I have mentioned above that half the spermatids receive an extranuclear body which is not found in the other half , and it is conceivable that this is connected with the difference under discussion .
But the fact that in the Bee half the spermatid nuclei degenerate , while in other related Hymenoptera all develop into spermatozoa , suggests that two kinds of spermatozoa may exist .
And a similar condition has been described by Morgan* in Phylloxera , and confirmed in a fuller paper by von Baehrf in Aphis .
They find that there is an odd number of chromosomes in the somatic nuclei of the male , and at the first division of the spermatocytes one chromosome goes over undivided to one end of the spindle , so that one daughter nucleus receives a " heterochromosome , " while the other does not .
In the second division the heterochromosome divides equally , but only those secondary spermatocytes which contain it develop further , the others degenerating .
Thus , as in the Bee , half the daughter-cells of the primary spermatocytes atrophy , and since all fertilised oggs become females , Morgan supposes that the heterochromosome , which is contained in the functional spermatids , bears the female determinant .
A different explanation of the facts will be offered below , but the facts themselves clearly indicate the existence of two kinds of spermatids , one of which , as in Neuroterus , contains a body lacking in the other .
We may suppose , then , that in Neuroterus , one kind of spermatozoon , fertilising an egg of the summer generation , gives rise to female-producing ( thelytokous ) female , the other kind to a rnale* 'Proc .
Soc. Exper .
Biol , and Med. , ' 1908 , vol. 5 , p. 56 .
( Full paper , ' Journ. Exp. Zoo .
, ' vol. 7 , 1909 , p. 239 .
) t ' Zool. Anzeiger , ' vol. 33 , 1908 , No. 15 .
1909 .
] Gametogenesisof the , etc. producing ( arrhenotokous ) female .
In the Bee , as far as is known , all queens produce similar eggs , and this would be accounted for by the degeneration of one kind of spermatozoa .
The sex-determination in the Gall-fly would then take place as follows:\#151 ; The spermatocytes of the male would bear a determinant for maleness , which may be represented by the symbol $ .
At the nuclear division this passes into one daughter-cell , its fellow being without it and thus containing no sex-determinant , a condition which may be represented by the symbol \#169 ; .
The fertilisable eggs after maturation are supposed all to bear one determinant for femaleness , represented by the symbol ? .
The zygotes all develop into parthenogenetic females , but since half of them were fertilised by $ -bearing spermatozoa , and half by spermatozoa with no sex-determinant ( \#169 ; ) , the former will have the constitution ?
$ , the latter $ \#169 ; .
The females with constitution ?
\lt ; $ lay eggs which undergo maturation ; the ?
determinant is expelled , and the egg left with the \lt ; $ determinant becomes a male .
The females with constitution ?
\#169 ; lay eggs which undergo no reduction , but containing the ?
determinant develop into females .
The ?
determinant is transmitted to the eggs of the summer generation , and since only one sex-determinant is present in the egg , it remains in the pronucleus when the polar mitoses take place , with the result that all eggs of the summer generation are ?
-bearing , as is assumed above .
It now remains to be seen whether this scheme is consistent with what is known of sex-determination , first in ordinary cases where all eggs are fertilised and secondly in the Bee and other aculeate Hymenoptera .
The ordinary bisexual cases will be considered first .
As the result of my work on heredity and sex-determination in the moth , Abraxas grossulariata*and of other similar cases , Bateson]* suggested the hypothesis that the female is heterozygous in respect of sex , containing male and female determinants , the male homozygous containing only male .
Eggs would thus be produced in equal numbers bearing maleness or femaleness , but all spermatozoa would bear the male determinant .
This hypothesis would completely explain the cases which led to its formulation , but it is not completely consistent with the existence in many insects of two kinds of spermatozoa , one containing a heterochromosome and one without it .
More recently , the evidence from some forms of sex-limited inheritance , such as colour-blindnessj and congenital nystagmus , S have made it clear that there * 1 Evol. Committee Roy .
Soc. , Report IV , ' 1908 , p. 53 .
+ ' Science , ' 1908 , N.S. , vol. 27 , p. 785 .
t Bateson , " Mendel 's Principles of Heredity , " * Camb .
Univ. Press , ' 1909 , 1st impression , p. 230 , inserted slip ; 2nd impression , 1909 , p. 195 , note .
S See Lloyd Owen , 'Ophthalmic Review , ' vol. 1 , 1882 , p. 239 .
Mr. L. Doncaster .
[ Sept , 21 , are not only two kinds of eggs differing in their sex-determinants , but also two kinds of spermatozoa .
In colour-blindness , for example , the affection is dominant in the male and recessive in the female ; so that heterozygous males exhibit it , heterozygous females do not ; but affected men do not transmit it to their sons .
The sons of affected males are normal , but the daughters of affected males transmit the disease to some of their sons and , through some of the daughters , to their grandsons .
It is clear , therefore , that among the spermatozoa of an affected male , those only bear the determiner for the disease which will give rise to female zygotes ; but among the ova of a female heterozygous for the colour-blind condition , the colour-blind factor is borne indiscriminately by about half the ova , whether male-producing or female-produci ng .
These different facts all fall into line if we assume that the female contains both female and male sex-determinants , and produces equal numbers of ova bearing each ; while the male contains no female determinant , but produces two kinds of spermatozoa , one bearing the male determinant , the other being without any determinant for either sex .
Using the same symbols as before , viz. , $ and \lt ; $ for the female and male sex-determinants respectively , and the symbol O for absence of sex-determinant in a gamete , females would then have the constitution ?
and would produce ?
and eggs in equal numbers , males would be \#169 ; , and produce spermatozoa $ and \#169 ; in equal numbers .
In the insects with heterochromosomes , in the male the hetero-chromosome is regarded as bearing \lt ; $ , its absence being represented by \#169 ; ; in the female the two heteroehromosomes bear ?
and respectively .
If we regard the sex-determinants as equivalent to Mendelian allelomorphs , they must be considered as two pairs , each determinant being allelomorphic with its absence , i.e. ?
with \#169 ; , and with \#169 ; , but ?
and are spuriously allelomorphic with each other , so that when they coexist in the same zygote they cannot both enter the same gamete ; just as in the case of Abraxas , the \#165 ; determinant and the grossulariata determinant exhibit spurious allelomorphism , and are never found in the same gamete .
We shall then have females producing equal numbers of $ eggs and \lt ; $ eggs , males producing equal numbers of and \#169 ; spermatozoa , and it must further be assumed that ?
eggs are fertilised by spermatozoa , giving females ( $\lt ; ?
) , \lt ; ?
eggs by \#169 ; spermatozoa , giving males ( $ \#169 ; ) .
This scheme is consistent with the facts in three categories of cases which have hitherto seemed irreconcilable , viz. , Abraxas and those cases which resemble it : the insects with heterochromosomes , and the cases of sex-limited inheritance of which colour-blindness is the type .
In Abraxas there is spurious allelomorphism between the ?
determinant and the grossulariata ^909 .
] Gametogenesis of the Gall-Fly , etc. factor , so that ?
eggs are without that character ; $ eggs have it , but in the male , since the ?
determinant is absent , there is no such spurious allelomorphism , and the grossulariatafactor is indiscriminately distributed among the spermatozoa .
In insects with heterochromosomes these may be regarded as bearing the S and \#165 ; factors as suggested above .
In colour-blindness , the factor for the disease can only be borne by a gamete which contains a sex-determinant , but the latter may be either $ or ?
; in the male , the factor is borne by the S -bearing spermatozoa , but not by those which have no sex-determinant ( O ) ; since the latter fertilise $ -bearing ova and the former \#165 ; -bearing , an affected male transmits the factor for the disease only to his daughters .
But in the heterozygous female the colour-blind factor may be associated with either or ?
eggs , and thus some of her children of each sex receive it .
The suggestion may here be made in passing that this hypothesis of sex-transmission may explain the cases not rarely met with , especially in the offspring of hybrids , in which all or nearly all the offspring are of one sex .
If a $ egg were occasionally fertilised by a \#169 ; spermatozoon , the resulting individual would be a female containing no male determinant , and either all its eggs would be \#165 ; -bearing , or if any contained no sex-determinant they might be sterile or unfertilised .
In crosses which give only male offspring ( as in the case of the moths , Tephrosiacrepuscularia and T. described by Tutt* ) , the \#165 ; -bearing eggs may fail to attract the spermatozoa of the other species , and so only male offspring would result .
After this digression , we must now return to the case of the Bee , Wasp , and other parthenogenetic cases .
In a previous paperf I have made the suggestion ( which I find has also been made by Morgan ) that the presence or absence of a spermatozoon in the egg might cause the maturation to take place differently .
It is possible that when a spermatozoon is present the egg of the Bee undergoes normal " reduction " divisions , halving the chromosomes both quantitatively and qualitatively , and removing the $ determinant .
But when the egg develops unfertilised the polar divisions may both be equational , leaving the haploid ( reduced ) number of bivalent chromosomes in the egg , and in this case expelling the \#165 ; determinant .
This would be comparable with the supposed male eggs in the spring brood of and in each case in the germ-cells of the male the reduced number would remain throughout , but in the body-cells the bivalent chromosomes would separate * Tutt , 'Trans .
Ent .
Soc. , ' 1898 , Part I , p. 17 .
Quoted by Castle , " Heredity of Sex , " ' Bull .
Mus. Comp. Zoo .
Harvard , ' vol. 40 , 1903 , p. 206 .
t " On the Maturation of the Egg in the Teuthredinidae , " ' Q.J.M.S. , 'vol .
49 , 1906 , p. 586 .
Mr. L. Doncaster .
[ Sept. 21 , into their constituents , so giving the same number in male and female body-cells , as is observed .
In the male Bee a single $ determinant would be present in the spermatocytes : this would pass into one spermatid , the other being left with none ( \#169 ; ) .
The latter spermatids are the " male polar bodies " which degenerate .
All the spermatozoa thus contain the factor , their presence in the egg causes the $ factor to remain in the polar divisions and the fertilised egg is thus ?
$ and becomes a female .
This assumption that the presence of a spermatozoon in the egg causes a difference in the maturation process , and leads to the removal of the male instead of the female sex-determinant seems improbable at first , sight , but there are considerations which reduce its improbability on further examination .
If both eggs and spermatozoa are commonly of two kinds , the hypothesis of selective fertilisation cannot be avoided , and if there is some definite attraction between a ^-bearing and $ -bearing nucleus , the presence of a $-bearing spermatozoon ( or male pronucleus ) in the egg may attract the ?
determinant to the inner end of the maturation spindle , although in an egg with no spermatozoon it would be expelled .
Clearly some such attractive force exists , the nature of which is at present unknown , for otherwise in the saw-flies the male pronucleus might conjugate equally often with one of the polar nuclei , which to all appearance exactly resemble the egg-nucleus , instead of with the egg-nucleus itself .
If it is assumed that , in cases where the egg matures before the entrance of the spermatozoon , the sex-determinant remaining in the egg attracts only one of the two kinds of spermatozoa , it does not seem a very improbable extension of the hypothesis that the presence of one kind of spermatozoon in the unmatured egg may cause the expulsion of the corresponding sex-determinant at the maturation division .
In Wasps and Ants* the spermatogenesis takes place much as in the Bee , except that two apparently similar spermatids are formed by the spermatocyte division , each of which becomes a spermatozoon .
Their spermatogenesis thus closely resembles that of Neuroterus , and the most probable assumption seems to be that two kinds of spermatozoa are formed , $ and \#169 ; respectively , but that only the \lt ; $ -bearing are functional .
They would thus be exactly comparable with the Bee , except that the reduction of the useless spermatozoa is not carried so far .
Other possibilities are of course not excluded , e.g. , that the d determinant divides and passes into both spermatids , or that , as in Neuroterus , functional \#166 ; $ and 0 spermatozoa occur , and therefore that among the offspring of the female which founds the nest there may be two kinds of female individuals produced , one of which would contain no $ determinant .
In ants and wasps the workers are not sharply separated from the queens * Meves and Duesberg , 'Arch .
Mikr .
Anat .
, ' vol. 71 , 1908 , p. 572 .
1909 .
] Gametogenesis of the etc. 109 as they are in the Bee , and these purely female individuals might develop into workers which would not lay eggs .
It is generally considered that when worker ants lay eggs , these always yield males ( as in the Wasp and Bee ) , but Keichenbach* describes the case of a nest in which the workers produced females except at the season when males normally occur , and then males were produced .
If some of the workers were of the ?
O constitution , others $ this might thus be explained .
The case clearly requires further investigation .
The hypothesis of sex-determination outlined above is doubtless highly speculative , but it has the advantage that it brings into line several sets of facts which have hitherto seemed irreconcilable , viz. , the results obtained by breeding such cases as Abraxas , and also such cases as colour blindness , the cytological observations of Wilson and others on heterochromosomes , and the peculiar behaviour of the Bee and other Hymenoptera which have facultative parthenogenesis .
The fact that in Neuroterus the parthenogenetic generation is separated from the bisexual one , and that some parthenogenetic females give rise only to males , others only to females , supplies the clue , for it is clear that some difference must exist in the constitution of the spring-brood females to account for the difference in sex in the offspring of different individuals .
That this factor is introduced by the male parent seems probable from the known fact that in so many cases spermatozoa of two kinds are produced .
In the Aphides a variety of conditions occurs in different cases , but here also the suggested explanation holds good .
In all species fertilised eggs yield females , and these produce a varying number of parthogenetic generations .
All these must be regarded as of constitution $ In some , the later parthenogenetic females give rise to both sexual females and males from the same individual ; f in these it may be assumed that in the eggs which develop into sexual females the \lt ; $ determinant is removed ; from those which yield males the $ determinant .
In other cases the parthenogenetic females give rise to " sexuparee , " some of which produce parthenogenetically only males , others sexual females .
In these , the female-producing sexuparse may have lost the \lt ; $ determinant , while the male-producing sexuparse contain both ^ and $ , but the ?
determinant is removed with the polar body of the egg which yields a male .
Morgan and von Baehr have shown in a species of Phylloxera and in Aphis saliceti that half the secondary spermatocytes degenerate , as do half the spermatids in the Bee , and in this case they have * ' Biol. Centralblatt , ' vol. 22 , p. 461 .
+ N. M. Stevens , 'Carnegie Inst. Public .
, Washington , ' No. 51 , 1906 ; also von Baehr , loc. cit. VOL. LXXXII.\#151 ; B. K Mr. L. Doncaster .
[ Sept. 21 , shown that those which degenerate lack the heterochromosome , i.e. those which we have represented by \#169 ; .
The spermatozoa are thus all -bearing , but since the sexual female lacks the $ element , all the eggs are $ , and the resulting offspring ( zygotes ) are females with constitution $ $ .
Morgan assumes that the spermatozoon determines the sex , but this involves the belief that the male Aphid contains a $ determinant , and since we know that the parthenogenetic female contains the factor ( since males are produced from it ) , this would involve the complex assumption of alternative sex-dominance .
If my hypothesis is correct , it could be tested in Aphis by observing whether , in the single polar division by which a male egg is produced , one complete heterochromosome is extruded , thus leaving only one heterochromosome in the egg instead of the two which are characteristic of the parthenogenetic females .
In this discussion I have not attempted to deal with the case of the saw-flies ( Tenthredinidse ) , in which facultative parthenogenesis occurs .
In former papers* I have discussed this question , but subsequent observations have led me to believe that the work requires extension and revision , and since this is still in progress , I prefer to leave any discussion until the results are clear .
A final word should be given to the recent hypothesis of Wilsonf and Castle , J that the sex-determinants are not - and ?
-bearing respectively , but that only one kind of determinant exists , and that the female contains one more such determinant than the male .
Castle supposes that in some species the female contains two such determinants , the male one ; in other species* the female one and the male none .
The existence of two kinds of parthenogenetic females in Neuroterus cannot be explained on either assumption , but it is conceivable that both kinds of parthenogenetic female contain one sex-factor and the male none .
The difference between the arrhenotokous and thelytokous female would then be that the eggs of the former contained a mechanism for expelling the sex-factor , while the thelytokous egg would retain it .
The extranuclear body found in half the spermatids might be responsible for this difference .
This , however , does not seem very probable .
The hypothesis , also , cannot explain the almost invariable excess of affected males in cases of sex-limited inheritance , such as colour-blindness , which is accounted for if we assume that the factor for the disease is more often borne by $ than by ?
ova .
In general , the hypothesis does not seem to bring together such a wide range of phenomena as that suggested above .
* ' Q.J.M.S. , ' vol. 49 , 1906 , p. 561 and vol. 51 , 1907 , p. 101 .
t ' Science , ' January , 1909 , vol. 29 , p. 53 .
| 'Science , ' March , 1909 , vol. 29 , p. 395 .
See also Morgan , 'Journ .
Exp. Zoo.* Sept. , 1909 , p. 332 .
Ill I go9.| Gametogenesis of the , etc. In conclusion , I wish to acknowledge the assistance received in the later part of my work from a grant from the " Endowment of Research Fund " of the Birmingham Natural History and Philosophical Society , and especially to record my indebtedness to Prof. Bateson , who has read the MS .
of the paper and given valuable help in criticisms and suggestions .
EXPLANATION OF PLATES .
The figures , with the exception of Nos. 20 and 22 , were drawn with Zeiss apochromatic objective , 3 mm. ( apert .
1 AO ) , and Zeiss ocular 12 , giving a magnification of about 1600 diameters .
When the magnification is less , it is mentioned in the description of the figure .
The outlines of the larger figures were drawn with camera lucida , but the chromosome groups are drawn free-hand .
Plate 1 .
Fig. 1.\#151 ; Chromosome group from anaphase of mitosis in developing wing of old larva , male .
Fig. 2.\#151 ; Two similar groups from female pupa , summer generation .
Fig. 3.\#151 ; Two similar groups ( the two ends of one mitotic figure ) from a pupa of the spring generation .
Fig. 4 a , b , c.\#151 ; Three anaphases , two in face and one side-view , of spermatogonial mitoses , half-grown larva .
Fig. 5 a , b.\#151 ; Two primary spermatocytes , after the growth period .
Male pupa .
Fig. 6.\#151 ; Prophase of first spermatocyte mitosis .
Chromosomes appear as about 10 irregular bands .
Fig. 7 a.\#151 ; Early stage of first ( abortive ) spermatocyte division .
A centriole is seen at each end of the cell , and a small dot on the upper edge of the nucleus .
Fig. 7 b.\#151 ; Later stage : the nucleus is re-forming , and the cytoplasmic bud with the centriole is being separated .
Fig. 7 c.\#151 ; Rather earlier stage : pear-shaped nucleus and finger-like process with centriole at its tip .
Fig. 8 a , b , c.\#151 ; Secondary spermatocytes , three stages of re-formation of the nucleus. .
In 8 band c the cytoplasmic bud at one end , and the stained dot near the nucleus , are visible .
Fig. 9.\#151 ; Early propliase of second mitosis .
More than 10 chromatin bodies are visible , although part of the nucleus is not included in the section .
At this stage the chromatin bodies are not always so definite as in this case .
Fig. 10 a , 6 , c.\#151 ; Three prophases of second spermatocyte mitosis , showing 10 band-like chromosomes .
In 10 a not all the chromosomes are represented ; in 10 c the little knob on the lower side of the cell perhaps represents the " cytoplasmic bud .
" Figs. 11 and 12.\#151 ; Equatorial plates of second mitosis ; in 12 the stained dot is seen outside the circle of chromosomes .
Fig. 13 a , b.\#151 ; Two metaphases in side view .
Fig. 14 a , b.\#151 ; Anaphases in side view .
In each the stained dot is seen near the lower end of the spindle .
Fig. 15 a , b.\#151 ; Telophases : stained dot seen included in one daughter-cell .
Fig. 16.\#151 ; Two telophases seen in pole-view .
Fig. 17.\#151 ; Two early spermatids .
K 2 112 Gametogenesis of the Gall-Fly , etc. Fig. 18 a , b.\#151 ; Two prophases from a probably abnormal follicle .
In 18 a the double number of chromosomes is seen , in 18 b they appear to be pairing longitudinally .
In 18 bnot all the chromosomes are shown .
Fig. 19 a , b. Anaphases ^from mitoses in ovary of half-grown female larva , summer generation .
a , polar view ; b , side view .
Fig. 20.\#151 ; Part of egg tube from pupa of summer generation , showing alternation of developing egg on right , with undifferentiated primitive ova on left .
( x about 650 .
) Plate 2 .
Fig. 21 a , b , c.\#151 ; Nucleus of developing egg of spring generation , a and b are optical sections of the nucleus at different levels , c part of the same nucleus in the next section of the series .
In b the chromosomes shown in outline are ends of those seen in the same positions in a. Figs. 22 to 33 represent the maturation of the summer egg .
Fig. 22.\#151 ; Outline of section of egg showing stalk and polar mitosis .
( x about 650 .
) Fig. 23.\#151 ; Nucleus at edge preparing for maturation division .
Fig. 24 a , b.\#151 ; -Two successive sections of early stage of division .
Fig. 25 a , b.\#151 ; Polar views of first maturation division , a represents the inner , b the outer chromosome group .
Drawn from different levels in the same section .
Fig. 27 represents nearly the same stage in side view .
Fig. 26 a , b , c.\#151 ; Later stage of maturation , a and b are drawn from different levels in the same section , c from the next section of the series , a represents the innermost group , and some of the chromosomes in b are apparently continuous with those in a. c represents the outermost group .
Fig. 27.\#151 ; Early stage of the first maturation division , side view .
Fig. 28 a , b , c , d.\#151 ; a and b are the two groups of polar chromosomes drawn from the same section at different levels , c represents the spermatozoon becoming the male pronucleus in the next section ; d , the insinking chromosomes forming the female pronucleus cut twice in the succeeding two sections .
Fig. 29.\#151 ; Early stage of second maturation division , unusual condition .
Fig. 30 a , b.\#151 ; Close of first and beginning of second maturation divisions ; two successive sections of series , a shows outer group , b part of cuter and elongated inner group .
Fig. 31.\#151 ; End of second division of inner group .
Chromosomes which form female pronucleus closely packed , below ; inner polar chromosomes and outer group preparing to divide , above .
Female pronucleus in next section to polar chromosomes .
( x about 800 .
) Fig. 32._Close of second maturation division .
Chromosomes of female pronucleus sinking in , outer group completing its division , represents the spermatozoon becoming the male pronucleus , from the next section .
( x 800 .
) Fig. 33.\#151 ; Female pronucleus ( pr .
n. ) and three groups of polar chromosomes .
Plate 3 .
Fig. 34.\#151 ; Meeting of male and female pronuclei .
Fig. 35.\#151 ; Conjugation of pronuclei and first segmentation spindle , a and b mark the chromosome groups derived from the two pronuclei .
In the section they lie at different levels , so that they are actually placed across the axis of the spindle .
' Fig. 36 a , b._Two polar views of telophases of segmentation mitoses , showing about 20 chromatin masses .
Side view of same stage in fig. 38 .
Lancaster .
Fig. 1 \ * .* .
V \gt ; Fig. 2 * # v* i ?
ot/ .
Soc. Proc. B , vol. Plate 1 .
Fig. 4 Fig. 3 # 0 Fig. 5 M*\#171 ; jjS Fig. 6 SSS*i\gt ; riff V Fig. 7a e to\ Fi\#163 ; .
7\#163 ; v*f . .
Fig. 7c a I .
ft*I XL/ r \ % o I Fig. S mJot : S6 8a sff : v KJ 7 .
% \lt ; 1 % F^ .
72 X / Srv^ Fig. / 5a 3- * Ft# .
15b Fig. 9 Fig. 10 \#163 ; \gt ; * .* ' T^ \#165 ; Sfe# m Xv f/ \gt ; V/ 0\#171 ; 10b 10c % Fig. / r Fig. 14a \amp ; Fig. 74\amp ; Fig. / 5 $ ?
' .
" - -T / ( / l^ ' W/ *\#187 ; i fn\gt ; / 13a 13b Fig. 16 Fig. 17 Fig. / g fJMi Si ' ' '\#166 ; '0 0 \lt ; **\gt ; vji.l .n\gt ; \#171 ; .1\gt ; 3 18a 18b Fig. 19 Fig. 20 t / 9a 19b London Stereoscopic Co. imp .
\Doncaster .
Roy. .
So .
Proe .
B , vol. Plate 2 .
Fig. 21 Fig. 32 Fig. 33 London .
Stereoscopic Co. imp .
Doncaster .
Roy .
Soc. Proc. B , vol. Plate 3 .
Fig .
35 Fig. 43 Cell Lamination of the Cerebral Cortex of , etc. 113 Figs. 37 and 38.\#151 ; Two stages of segmentation mitoses .
Figs. 39 to 48 represent eggs of the spring generation .
Fig. 39 a , b.\#151 ; Two successive sections of chromosome group at edge of egg before maturation division .
Fig. 40.\#151 ; Early stage of maturation division .
Fig. 41.\#151 ; Late stage of maturation division .
Fig. 42.\#151 ; Completion of maturation division .
Chromosomes of egg-nucleus sinking in ; polar chromosomes in two groups .
Fig. 43 a , b , c.\#151 ; Polar chromosomes and egg-nucleus , a and b , successive sections of polar chromosomes ; c , egg nucleus several sections removed .
( x about 800 .
) Fig. 44 a , b.\#151 ; Two equatorial plates of segmentation mitoses , showing about 20 chromosomes .
Fig. 45.\#151 ; Prophase of segmentation mitosis , showing numerous long coiled chromosomes .
Fig. 46 a , b , c.\#151 ; Metaphase ( a ) , early ( b ) , and late anaphases ( c ) of normal segmentation mitoses with diploid number .
Fig. 47.\#151 ; Equatorial plate in face of segmentation mitosis with haploid number .
Fig. 48 a , b , c.\#151 ; Metaphase ( a ) , anaphase ( b ) in side view , and anaphase in pole view ( e ) of segmentation mitoses with haploid number .
Preliminary Note upon the Cell Lamination of the Cerebral Cortex of Echidna , with an Enumeration of the Fibres in the Cranial Nerves .
By Edgar Schuster , D.Sc .
, Fellow of New College ( Pathological Laboratory , Claybury Asylum , Essex , and Department of Comparative Anatomy , University Museum , Oxford ) .
( Communicated by Dr. F. W. Mott , F.R.S. Received September 30 , \#151 ; Read December 9 , 1909 .
) [ Plates 4 and 5 .
] Material.\#151 ; The following notes are based on the study of the brain of an Echidna which died in the gardens of the Zoological Society in London .
Dr. F. W. Mott , F.R.S. , kindly placed the brain in my hands with the suggestion that I should examine the cell lamination of the cortex and should estimate the numbers of fibres in the cranial nerves .
For this and for his advice and help during the investigation I wish here to express my gratitude .
The right hemisphere was cut transversely into a series of sections 10 / i in thickness , from which sections were taken at intervals of about \#163 ; mm. , stained with polychrome blue , and mounted .
It may , perhaps , be mentioned

