598 



NA TURE 



[October 21, 1897 



discussion of all the available observations of this star up to the 

 year 1894 relatively to the form and changes of the light curve. 



Comet Perrine. — We have received a telegram from Prof. 

 Kreutz, Kiel, dated October 18, in which we are informed that 

 Comet Perrine on October 16, at gh. 38 'Sm., Lick mean time, 

 appeared of the eighth magnitude, and was situated in R.A. 

 3h. 36m. 8s. and N.P.D. 23° 13' 16". It was observed to have 

 a small tail. 



A later telegram, dated October 19, gives the following position 

 and magnitude: October 18, llh. 31 "im., Pola mean time, 

 R.A. 3h. 25m. 31S., N.P.D. 20° 34' 44", magnitude lO'o. 



HEREDITARY COLOUR IN HORSES. 



TVyr Y attention has been drawn to a collection of data on the 

 ■^ hereditary transmission of colour in horses, which appeared 

 in the last Christmas Number of the Horseman, a newspaper 

 published in Chicago, U.S. It is signed with the pseudonym 

 of " Tron Kirk." I corresponded with the author, who is noted 

 for his knowledge of horse-breeding, and he assures me of their 

 substantial correctness. His statistics are chiefly obtained from 

 breeders' catalogues, and, however valuable in other ways, 

 fail seriously through the great disproportion which must exist 

 between the number of the different sires and that of the 

 dams, a single sire in the polygamatous arrangements of a 

 .stud begetting a numerous offspring from nearly as many 

 dams. It is stated that no less than 3100 foals were begotten 

 by only 46 difTerent bay sires, or more than an average of 67 

 foals by each sire. Now the number of offspring of the 16 

 different forms of colour union registered in Table I. is, with 

 one exception, by no means large ; in 9 cases it is less than 100, 

 and in one of these it is only 6. Consequently the prepotencies, 

 or the reverse, of individual sire will fail to balance each other, 

 and are sure to produce anomalous results. 



The data I propose to use are those contained in Table I. ; 

 they have been extracted from the memoir in the Horseman, 

 but are newly arranged both in line and column. I have 

 omitted grey altogether, no grey stallions being recorded, and 

 all the grey foals coming from grey dams. 



Table I. 



My first inquiry was to determine whether the sire or the dam 

 exercises the larger influence in transmitting his or her own 

 colour to the offspring, this being a point on which different 

 breeders express contradictory opinions. The truth in the 



NO. 1460, VOL. 56] 



present instance is easily arrived at by means of Table 11.^ 

 where the percentages of the offspring who resemble their D.am are 

 compared with those that resemble their Sire. This is done in 

 each several pair of " reciprocal " unions, such as that which 

 consists of \Dam, bay — Sire, brown]; and of [Dam, brown — 

 Sire, bay]. The table shows a total of 394 cases of resemblance 



Table II. 



in the one set, to 381 in the other ; in short, it proves that the 

 potency of the dam in tran.smitting colour is substantially the 

 same as that of the sire. 



The intention of the second inquiry was to test an important 

 part of my recent theory on "The average contribution of 

 each several ancestor to the total heritage of the offspring *' 

 (Pror. R. Soc, June 3, 1897, and Nature, July 8, 1897). 

 According to this theory each of the two parents contributes 

 on the average one quarter of the total heritage, each of 

 the four grandparents one sixteenth, and so on. If this be 

 strictly true, and if the potency of the two sexes be the 

 same, one half of the varied offspring from the [bay — bay] 

 pairs added to one half of those of an equal number from the 

 [brown — brown] pairs, would be identical in character with the 

 same number of the offspring of [bay — brown], also with those 

 of [brown — bay]. The same holds true for every other form of 

 union between sires and dams of different colours. However, 

 the statistics in Table I. run so roughly that this particular com- 

 parison would fail to lead to trustworthy results. It is true that 

 reciprocal unions are seen to give rise to similar results in 

 [chestnut and bay], to fairly similar ones in both [bay and 

 brown], and in [brown and black], and to not very dissimilar 

 ones in [bay and black], but each of the two remaining sets is. 

 incongruous. Moreover, the figures contained in them run 

 wildly ; thus in the line [black— chestnut] the sequence of the 

 numbers, 30, 40, o, 30, is a statistical impossibility, and in the 

 line [chestnut — brown] the sequence of o, 86, 11, 2, is very 

 suspicious. It is obvious that a more trustworthy interpretation 

 of the true state of the case might be deduced from these rude 

 data, if the four entries in each line could be appropriately con- 

 solidated so as to be expressed by a single number. It occurred 

 to me that a good way of doing so would be to determine the 

 amount of red pigment corresponding to each entry in the same 

 line, and to sum those amounts. Guided at first by the judg- 

 ment of the eye, and afterwards by gbserving how nearly each 

 successive assumption satisfied the observed facts, I fixed on the 

 following allowances, supposing full red pigmentation to count 

 as I. For chestnut, O'S ; for bay, 07 ; for brown, 04; and, 

 recollecting the considerable amount of red pigment in the 

 blackest hitman hair, I fixed the allowance for black at 01. 



There are twelve equations in which these four values appear; 

 so if all are fairly well satisfied by the above assumptions, we 

 may rest content. I did not take pains to have the red pigment 



