394 Prof. W. F. R. Weldon .
on the [ Jan. 15 , Scawen Blunt* for the Kehailan strain of Arabs .
The same assertion is also made for the horses called " rote " ( ?
red chestnut ) in the Jutland breed by Jensen.f Finally , we have a general statement by Hayes , !
applicable to all kinds of horses , that in the vast majority of cases a foal from a chestnut dam by a chestnut sire is of a chestnut colour .
The existence of the two kinds of dominants , and the conclusions based thereon have not , so far as I know , been previously recognised .
Note on the Offspring of Thoroughbred Chestnut Mares .
By W. F. R. Weldon , F.B.S. ( Received January 15 , \#151 ; Read January 18 , 1906 .
) The colours of English thoroughbred race-horses , as recorded in Weatherby 's General Stud-Book , are grouped under the six main categories\#151 ; Grey , Roan , Chestnut , Bay , Brown , Black ; but each category includes a considerable range of colour , and intermediates occur with quite sensible frequency .
At the meeting of the Royal Society , held on December 7 last , the suggestion was made that the relation between the most important colours , Chestnut , Bay , and Brown , might be expressed by a simple Mendelian formula\#151 ; Bay or Brown being regarded as determined by one Mendelian unit , which was " dominant " to the " recessive " Chestnut .
In discussing this suggestion , !
made certain statements , which I promised to justify as soon as possible , by offering the data , on which they were based , for publication .
The object of the present note is partly to fulfil my promise , and partly to call attention to certain facts which must be considered in the attempt to apply any Mendelian formula whatever to the inheritance of coat colour in racehorses .
I have taken from Weatherby 's Stud-Book a fairly complete record of all the foals produced in England by Chestnut mares during the period ( rather more than eight years ) covered by vols .
18 and 19 .
All cases of doubtful paternity have been excluded , and I have not tried to obtain external evidence in the few cases where the colour of sire or of foal is omitted .
With * 1 The Nineteenth Century and After , ' 1906 , January , p. 63 .
t ' Deut .
Pferdezucht .
, ' I , Hft .
11 , 1904 .
( Original not seen ; abstract in ' J. B. Landw .
Pfl .
u. Tierziichtung , ' 1905 , II , p. 273 .
) | ' Points of the Horse ' ( 3rd ed. ) , 1904 , p. 326 .
1906 .
] Offspring of Thoroughbred Chestnut Mares .
395 these limitations , I believe the following statements are substantially complete and accurate .
The Chestnut mares recorded produced , in the period dealt with , 5643 foals by 726 sires , the colours of the sires being as follows :\#151 ; Colour of sire .
No. of sires .
( Colour of sire .
No. of sires .
Ch. 214 Br .
or Bl .
10 Ch. or Ro .
2 Bl .
8 Ch. or Bl .
1 Gr .
3 B. or Ch. 1 Bo .
or Gr .
1 B. 340 Br .
or Gr .
1 B. or Br .
24 Gr .
or Bl .
1 Br .
120 726 1 It will be seen that 41 of these sires are intermediate between some two of the accepted colour-categories .
Omitting the Eoan or Grey sires , the foals produced are shown in the following table :\#151 ; 1 Offspring of Chestnut Mares .
Colour of sire .
Colour and frequency of foals .
Ch. \#171 ; Int. Ch. B. B. or Br .
Br .
Br .
or Bl .
Bl .
Totals .
Ch 1568 18 2 3 1 2 1594 B 1118 7 1449 89 132 4 8 2807 B. or Br 44 1 85 9 21 2 1 163 Br 233 2 439 44 146 13 14 891 Br .
or Bl 22 \#151 ; 5 1 5 1 3 37 Bl 13 \#151 ; 28 5 11 2 8 67 2996 10 J 2024 150 318 23 36 | 5559 1 .
Chestnuts Mated Together.\#151 ; It is clear that Chestnut cannot be regarded as a Mendelian recessive , since Bay , Brown , Black , and their intermediates , may be produced by the union of two Chestnut individuals .
The number of foals not Chestnut , resulting from such unions , is .
26 , or about 1 in every 60 of the whole number observed .
They were produced by 21 out of the 214 sires recorded , or about 1 sire in every 10 produced at least 1 foal which was not Chestnut .
Since the mean number of foals per sire is only 1594/ 214 = 7'4 through the period observed , the most probable number of Prof. W. F. R. Weldon .
on the [ Jan. 15 , sires giving foals of colour other than Chestnut , if each sire were capable of giving 1 such foal in 60 , would be 7'4/ 60 x 214 , or about 26 , the standard deviation of the expectation being nearly 4'6 ; so that the result obtained is not inconsistent with the view that any Chestnut horse is capable of producing , by union with a Chestnut mare , a small proportion of foals of some colour not Chestnut .
If each sire had produced but a single foal , we should clearly have to expect that only 1 sire in 60 would produce a foal of abnormal colour .
Among the sires recorded , 56 have produced only a single foal each , and of these foals one is Bay or Brown .
On the other hand the Chestnut sires , and Despair , have produced 115 foals , all Chestnut , while Amalfi , Brag , and Dog Rose have together produced 36 Chestnuts and seven of other colours during the period examined .
We may conclude : ( 1 ) That the mean chance of obtaining a foal which is not Chestnut from two Chestnut parents is about 1/ 60 ; and ( 2 ) that this chance is not the same for all sires ; it is probably less than 1/ 60 for Amphion , Despair , and their like , and much greater for such sires as Amalfi , Brag , and Dog Rose .
This variability in the power of transmitting colours other than their own , possessed by Chestnut sires , is precisely what Galton 's theory of dominance would lead us to expect ; it is , however , a difficulty in the way of any Mendelian theory involving a reasonably small number of elements .
2 .
Chestnuts Mated with Bays and Browns.\#151 ; On the view that Bay is a Mendelian " dominant " to Chestnut , Bay sires should be sharply divided into two series : those which are " hybrid , " or of constitution " I)K , " giving 50 per cent , of Chestnut foals and 50 per cent , of Bays by the unions considered , and those which are " pure dominants , " giving only Bay offspring .
Any Bay horse with one Chestnut parent must be " hybrid " on this view ; but it does not follow because both parents are Bay that a Bay colt is a pure dominant .
In the table below , those Bay sires which had either one Chestnut parent , or were known to have produced a Chestnut foal , are classed as " DR " ; while a Bay whose parents were both of some colour other than Chestnut was classed as DD unless I could find a record that it had produced a Chestnut foal .
Among the 65 sires classed as DD , no less than 37 have produced only one foal each during the period dealt with .
Some of the sires are so young that their whole progeny from Chestnut mares is too small to afford a test of their real nature ; others are so old that the labour of testing them by reference to previous volumes of the Stud-Book is great , and involves risk of error .
For such reasons as these , the number of sires classed as DD is certainly rather 1906 .
] Offspring of Thoroughbred Chestnut Mares .
above than below the truth .
In spite of this , however , the proportion of Chestnut foals produced by the sires called DR is too small for any simple Mendelian theory .
The simple Mendelian expectation is half , or 1235 Chestnuts , the standard deviation being ^2470/ 4 , or about 24'8 .
The observed deviation is \#151 ; 77 , or more than $T times the standard deviation , and a reference to the Tables of the Probability Integral shows that the chance of getting a deviation so great as this is only 0-0009 , or a little less than one in a thousand .
Chestnut Mares x Bay Horses .
Colour and frequency of foals .
Sire .
Ch. Ch. Ro .
Ch. Bl .
Ch. B. B. Br .
Br .
Bl .
Bl .
Totals .
" DR " !
1118 1 1 4 1141 1 68 88 4 5 2430 " DD " \#151 ; \#151 ; 1 308 21 44 3 377 1118 1 1 5 1449 89 132 4 l 8 2807 The matings with Brown horses , treated in the same way , give similar results .
Chestnut Mares x Brown Horses .
Sire .
Colour and frequency of foals .
Ch. Ch. B. Ch. Br .
B. B. Br .
Br .
Br .
Bl .
Bl .
Totals , " DK , " 233 1 1 246 193 25 19 72 74 9 4 9 5 594 297 " DD " 233 1 1 439 44 1 146 13 14 891 i Here the simple Mendelian expectation is 297 Chestnuts , so that the observed deviation from the required result is \#151 ; 64 , the standard deviation of the expected result being y/ 594/ 4 = 12-2 , or the observed result is even more improbable than that given by Bay sires .
If we attempt to consider some more complex Mendelian theory of the relation between Chestnut and Bay or Brown , we must base our hypothesis of the number of gametic determinants involved upon some separation of 398 Note on the Offspring of Thoroughbred Chestnut Mares .
sires into groups , those of each group giving a definite proportion of Chestnut foals when mated with Chestnut mares .
My observations do not at present decide this point ; but through the period I have dealt with no such separation of sires into groups seems possible .
Taking only the Bay or Brown sires which have given at least 20 foals in the period covered by my records , I find_ Sire .
Foals .
Ch. Not .
Oberon 1 42 Blue Grreen 1 28 Minting 2 31 Y eracity 2 29 at one end of the series and at the other such sires as ( 21 Ch. + 10 Not ) , or Queen 's Birthday ( 13 Cli .
+ 3 Not ) , which offer a transition to such Chestnut sires as Amalfi .
The detailed discussion of individual sires , and of the proportion of Chestnuts yielded by each , must be postponed until the work of extracting the records from the Stud-Book has been completed .
In the meantime , however , the following points seem clear:\#151 ; 1 .
No simple Mendelian view of the relation between Chestnut , Bay , and Brown , regarding Chestnut as a simple recessive , can be maintained .
2 .
The chance of getting a Chestnut foal from a Chestnut mare is not constant for sires of any colour whatever , and there is no indication that sires of any colour can be sorted into groups such that those in each group will give Chestnut foals in a Mendelian proportion when mated with Chestnut mares .
3 .
These points , together with the values for parental and grand-parental correlations already given by Professor Pearson* make it probable that the facts of inheritance of coat colour in horses can be expressed in terms of the hypothesis outlined by Mr. Galton in 1872 , f and developed by him in his subsequent writings .
I hope to discuss the application of Mr. Galton 's hypothesis to these data before long ; but the extraction of pedigrees from the Stud-Book is an operation which takes time , and I have felt bound to offer this imperfect statement in justification of what I said in the discussion to which I have referred , without waiting tmtil my work should be completed .
* ' Phil. Trans. , ' A , vol. 195 .
t ' Roy .
Soc. Proc. , ' No. 136 , pp. 394\#151 ; 402 .

