Physical Conformation of Cows and Milk Yield 



J. Arthur Harris 

 Station for Experimental Evolution, Cold Spring Harbor, N. Y. 



IN a quotation from the Board of 

 Agriculture on conformation of cows 

 and milk yield which recently ap- 

 peared in the Journal of Heredity, 

 (p. 253, June, 1915) it is pointed out 

 that so far as the studies of J. Reimers 

 go, they show that there is no correlation 

 between physical conformation and milk 

 yield. 



The results (at least as far as relation- 

 ships close enough to have any practical 

 significance for purposes of predicting 

 yield from conformation are concerned) 

 are quite what those who have an 

 extensive first hand acquaintance with 

 biometric work would have expected. 



The purpose of this note is to call 

 attention to and suggest an explanation 

 for an anomalous series of data of this 

 kind. G. Korreng has recently ex- 

 pressed the conviction (jahrbuch fur 

 wissenschaftliehe und praktische Tier- 

 zucht, vol. VII, pp. 132-142, 1912) 

 that there is an intimate negative 

 relation.ship between width of nether 

 jaw, "Ganaschenweite," and milk yield 

 — animals with the narrower jaws being 

 the best milkers. In substantiation 

 of his views he gives a series of 112 

 measurements and yields, which actually 

 show a strong negative correlation 

 between width and yield. The fre- 



quency of indi\'iduals having various 

 widths and yields is shown in the 

 accompanying table, which has been 

 drawn uj) from his data. 



In Figure 4, I have plotted out the 

 mean yields for animals of different 

 widths. They decrease with great reg- 

 ularity as width increases, as is shown 

 by the straight line fitted to the data. 

 Apparently, therefore, the data fully 

 substantiate the contention that there 

 is a relationship between conformation 

 and milk yield so close as to be of prac- 

 tical value in selecting the best milkers. 



If, however, one carries the analyses 

 of the records a step farther than 

 Korreng has done, and examines the 

 frequency distributions for the char- 

 acters under consideration as shown in 

 the totals of the table, and represented 

 in Figures 5 and 6, he sees at once 

 that both are distinctly bimodal. This 

 is particularly conspicuous in the case of 

 width. The two-humped condition is 

 not so pronounced in the case of milk 

 yield, but is nevertheless probably 

 significant. These conditions suggest 

 that Korreng's measurements were taken 

 on a group of animals that are not 

 racially homogeneous. A mixture of 

 heavy beef cattle giving a low milk 

 yield and light built dairy cattle would 



MILK YIELD IN LITERS 



348 



