194 Scientific Proceedings^ Royal Dublin Society. 



per cent. ; for we know grey to be dominant to the other colours (roan, 

 perhaps, excepted), and that, therefore, every grey foal must have at least 

 one parent grey; yet, among 248 grey foals noted in that volume, only nine 

 are entered as having neither parent grey. 



Inferences can be made with greatest confidence concerning greys. If a 

 dam whose colour is not given have a grey foal to a horse of another colour, 

 she is presumably grey ; if she have several sucli, she is grey without doubt. 

 As the colour descends towards chestnut, confidence also descends, and greater 

 care must be exercised in drawing conclusions ; but help can be got frequently 

 from the colours of parents and progeny, and from those of other near rela- 

 tives, even though the colours be well down the scale. For example, the 

 colour of the third parent (B 2) in the preceding table is not given in the 

 stud-book ; but the fact that six of her foals by five different sires were dun 

 is very strong evidence that she herself was dun; and this inference is con- 

 firmed, if that were necessary, by the further fact tliat one of her sisters and 

 both her brothers, all well-known animals, were of that colour. 



From the foregoing tables the position of dun can be worked out by more 

 than one method. We could employ the statistical method, and, by adding 

 up the total results of the matings of dun with other colours, find out which 

 colours it did and which colours it did not include. But, since the total 

 numbers are so small, it will be better, in the first instance, to find the position of 

 dun by considering the gametic composition of some of the best-known animals. 



The mother of this dun family, Silverlocks, was described as a chestnut. 

 Let us assume her coloui- to have been correctly described. She had three 

 dun foals to the Godolphin Arabian, who was a homozygous bay.' His 

 foals out of Silverlocks must, therefore, have got their dun colour from 

 their dam, in whom it must have been hidden by chestnut ; and, since each 

 of this homozygous bay sire's three dun foals must have carried a bay 



bay and brown (these two are not separated), and grey could have been made out easily and clearly. 

 Three of Dr. Crampe's conclusions as to what happens when both parents are of the same colour, 

 which are underlined in his summing-up, might be quoted ; — 



i. Chestnuts have chestnut foals exclusively, among which are chestnut-roans and chestnut- 

 greys (daruater stichelhaarige Fiichse und Fuchsschimmel). 

 ii. Blacks have both black and chestnut foals, and also roans and greys of these coloui's. 

 iii. Browns (that is browns and bays) have I'oals of all colours. 

 Crampe points out that the stud-books show exceptions to rule among chestnuts and blacks, but 

 that those exceptions, for the most part, are the result of erroneous entries : " der bei weitem grosste 

 Theil dieser Falle auf irrthiimlichen Eintraguugen beruht." Crampe did not discover the position of 

 roans and greys, or he would have seen why they are registered as occasionally occurring from other 

 coloui's. Koans and greys among browns caused him to say browns have foals of all colours. 



' The colours of at least fifty-six of his foals are known, and there was not a chestnut among 

 them. From chestnut mares he had thirteen bays, one black, and one brown; from bay mares, 

 sixteen bays and one brown ; from grey mares, seven greys and three bays ; and from mares of 

 unknown colour, one grey, one brown, and twelve bays. The browns and blacks are just such a pro- 

 portion as might be expected in view of tlie uncertainties in descriptions of bay, brown, and black. 



