]\gt ; To see whether symbiotic algae were present in the organs yielding enterochlorophyll , I examined fresh frozen sections , or portions ot the organ teased out in salt solution , but the results were negativo .
On steeping such preparations , first in alcohol , then in weak solution caustic soda , and neutrahsing with acetic acid , and afterwards testing with a solution of iodine in iodide of potassium and with Schultze'r fluid , I never obtained evidence of the presence of starch or celluloso .
Hence , apart from the absence of symbiotic algae under the microscope , this resuIt negatives their presence and also that of food chlorophyll .
The morphology of enterochlorophyll was studied in similar prepara .
tions , and on the whole it appears to be present dissolved in oil { globules and in granules , both of them enclosed in the epithelium lining the liver tubes .
It also occurs dissolved in the protoplasm oi the cells , and these appearances vary slightly in differenI cases .
It would therefore appear that enterochlorophyll is built up in the containing it ; that it is a chlorophyll , of which there arc several in animaIs , and that it is composed of two constituents , of which resembles closely the corresponding constituent of planl chlorophyll , while the other is generally slightly different , but that nc essential diiference exists between the respective pigments is proved by the fact that the constituents of both may be obtained crystallised in the same form .
In enterochlorophyll there is probably a more intimate union between the constituents than in plant chlorophyll .
All readings are reduced to wave-lengths , and the most important spectra mapped in the accompanying charts .
The appearance of enterochlorophyll under the microscope in different cases is also shown in the accompanying drawing , as as the crystals referred to above . .
" " Note on a Previous Paper By G. H. DARWIN .
F.R.S. , Fellow of Trinity College and Plumian Professor in the University of Cambridge .
Received March 19 , 1885 .
The paper entitled " " On the Stresses } in the Interior of Earth by the Weight of Continents and Mountains " " ( ' ' Phil. Trans.\ldquo ; Note on a previous Pape ' .
323 art I , 1882 , p. 187 ) has been found to be erroneous in certain points .
errors , however , do not touch the physical conclnsions there attained .
As this note has imporrance only in connexion with the aper , I proceed in the form of an appendix , without explanation of the notation .
In the first place\mdash ; Throughout the paper the normaI stresses require an addi ' term .
The only function of these stresses used in obtaining physical results is , and it remains unchanged when this correcis made .
The error its origin in S 1 .
Thomson 's solution ( 1 ) when reduced to the form applicable to the incompressible solid , is the solution of the equations , and two others .
The solution required is that of , and two othel'S .
The involved in my solution is not the potential of a true bodily force , but only an " " effective potential " " producing the same strains as those due to the weight of the continents and mountains , but causing a diflerent bydrostatic pressure .
When , therefore , is determined from Thomson 's solution , that is really equal to of the problem of the continents .
Hence equation ( 3 ) shouJd be , instead .
The correction to ( 3 ) must be carried on rough the rest of the paper , and obviously it merely adds to the stresses , leaving , P-Q , \mdash ; unchanged .
The error wonld have been avoided had I , as suggested on p. 190 , worked directly from the equations of equilibrium of the elastic incompressible solid , instead of from Thomson 's solution .
When the solid is compressible , this method of " " effective potenbial\ldquo ; ( see " " The Tides of a Viscous Spheroid " " Phil. Trans Part I , 1879 , pp. 7-9 ) for including all the effects of gravitation is not applicable without certain additional terms .
Hence in S 10 where the solid is treated as being compressible the expressions for the stresses are incomplete .
It will be found , however , that this incompleteness does not extend to the case of the mountains and valleys on the mean level surface , and that pol'tion of the section remains correct .
It ] not be difficult to make the requisite corrections bo the earlier part of the section , but I do not think it worth while to do so .
In the second place\mdash ; On p. 191 the following passage occurs:\mdash ; " " It may be seen from considerations of symmetry that if be a zonal harmonic , two of principal stress-axes lie in a meridional which he has pointed out and corrected .
It is clear that , in the limiting case of the zonal harmonic the surface becomes a series of parallel mountains and valleys on 1 flat surface , the principal stress parallel to the mountains must zero , and the above reasoning has led to a correct conclusion .
But in the case of the second zonal harmonic , with either excesa or deficiency of matter at the pole , there is a tendency for the equa torial regions to be either squeezed out or crushed in .
Now an out .
ward squeeze necessitates that the greatest pressure shall be perpen dicular to the meridian , and this is contrary to the general conclusion quoted above .
My error lay in overlooking this outward or inward tendency in the equatorial matter .
The conclusion is therefore wholly light in the case of the mountains and valleys , and at least partially wrong in the case of spheroidal deformation of the globe .
The data for examining into this question rigorously are given in my paper , and the besL way of treating the matter is to rewrite S5 on\mdash ; The State of Stress due to Ellipticity of Figure to Tide-generating Forces .
When the effective disturbing potential is a solid harmonic of the second degree , the soIution found will give the stresses caused by oblateness or prolateness of the spheroid .
It will also serve for the case of a rotating spheroid with more or less oblateness than is appropriate for the equilibrium figure .
When an elastic sphere is under the action of tide-generating forces , the disturbing potential dution will apply to this case also .
If we extract the case from Tables , and put in , and substitute colatitude for ]atitude , we have after some mple r [ Note the introduction of in the , in acoordance with first correction .
] Let , be the three principal stresses , each diminished by , so that\mdash ; ( ) .
Then\mdash ; .
Now let us find the surfaces , if any , over which or They are obviously given by\mdash ; \amp ; c This easily reduces to\mdash ; ( ) .
Thus the solutions are\mdash ; and and ( e ) By trial it is easy to see that at the centre and all along the polar axis , and that inside of the ellipsoid is greater than , and side it is less .
Hence inside of the ellipsoid and outside of it is the stress-difference .
nowhere vanishes so long as .
is not equal to , and vanishes where and , which is inside of the region for which is the stress .
This is the only point in the whole sphere for which the stress-difference vanishes .
The eUipsoid of separation cuts the sphere in colatitude and between the polar surface regions and the ellipsoid\mdash ; ( ) .
This Iast also holds for the whole polar axis , along which\mdash ; or .
[ In the paper the form ( g ) for was taken as applicable to the whole sphere ; the maximum value of the form ( g ) , and was therefore correctly computed .
] In order to find the actual value of in any special case , we to multiply the expression for by appropriate factors , determined in the paper .
For the present it will be convenient to omit these factors .
We may now from ( f ) and ( g ) determine the distribution of stress-diiference throughout the sphere .
By computation and graphical interpolation I have drawn th6 nnexed figure , showing the curves of equal stress-difference through .
urVes give the va , lues of , when the radius of the sphere is unity .
9 Me point marked is that where vanishes .
The dotted curve is the ellipse of separation ( e ) cutting the circle colatitude Over the polar cap and at the surface is constant and equal to ; at the surface from colat .
to the equator increases from 6 to 18 , varying as the square of the sine of the colatitude .
At the centre is 48 , being eight times the polar superficial value .
[ Beginning with the first sentence of p. 203 the remainder of S 5 .ill hold good .
It is well to observe , however , that where surface -difference is spoken of , it must be taken as to the caps only , the stress-diflerence at the equator being three times great .
It is worth while comparing the figurs 1 of the paper -$(Plate 19 ) with the figure now given .
We now come to the case of\mdash ; The Stresses due to the even Zonal Harmonics .
The complete ation of the regions within which : and are the proper measures of stress-diiference might be somewhat difficuIt .
As , however , these harmonics are only used for the determination of stress-difference in the equatorial regions , it is sufficient to find the boundary of the regions for that part of ) sphere .
We see from that only differs from by terms which depend on the square of the sine of the latitude .
Hence as far as the first power of we have Therefore if we neglect terms depending on the square of the sine of the latitude , we have from ( 22 ) , Then substituting , for , \amp ; c. , their values from ( 26 ) , and effecting some easy reductions , we find , .
VOL. XXXVIII .
whence Solving for the successive even values of , we find , when , as we already know .
the paper was always taken as being the stress-difference , and we now see that even when , the region is very thin in which this is untrue and where is the proper measure .
For the higher harmonics it soon becomes negligeable .
explains the transition from the incorrectness of the treatment in the paper of the case of the second harmonic f , o the correctness oi the treatment of the mountain ranges .
On looking at S7 and the accompanying figures we see that the ximum stress-difference occurs far within the region within which .
becomes the mean principal stress .
Thus S7 may be permitted to htand , save that in fig. 4 , Plate 19 , the ordinates of the .
curves , \amp ; c. , to be slightly mented at the surface where .
It is easy to see what small alterations are to be made in able V , and in the subsequent discussion , but clearly nothing material from a physical point of view need be amended .
It may be in conclusion that , whilst it is proper to correct mathematical errors in this paper , the ical c remain Iltouched .

