80 



EXPERIMENT STATION RECORI). 



and for period B $38.72; the net profit i>er cow per yejir for the different 

 periods was therefore as follows: Period A, $44. .39, period C, $r).'i.90; average 

 of periods A and C, $49.15; period B, $41.77, a difference of $7.38 per cow 

 aunnally, or 17.7 per cent in favor of the medium protein rations. 



" In spite of the fact that the high protein rations cost somewhat more, on 

 the average for the herd than those medium in protein, more milk and more 

 butter fat were produced on the latter and about 18 per cent larger net profit 

 per cow was obtained on the average for the respective periods of the 

 investigation." 



" While the general results of this investigation lead to the conclusion that 

 the exact amount of protein required in the feeding of dairy cows is less 

 important than was formerly believed to be the case, it is nevertheless certain 

 that a liberal supply of protein is necessary for the production of large amounts 

 of milk and butter fat ; it can be readily shown, in fact, that there is a direct 

 relation between the production of total solids or butter fat and the amounts of 

 digestible protein fed or the nutritive ratios of the rations fed dairy cows. . . . 

 The larger the production, the more feed the cows eat, and the narrower the 

 nutritive ratio becomes." 



" The teachings of the investigation described in this bulletin tend to empha- 

 size the importance of farm-grown feeds for dairy cows." 



Milk records, analyses, and cost of feeds, and an estimate of feed tequirements 

 for cows of different weights and productive capacity, are presented in tabular 

 form. 



The selection of feeds for dairy cows, F. W. Woll and G. C. Humphrey 

 {Wisconsin Sta. Bui. 200, pp. S-ll). — A brief and popular discussion of the 

 feeding of dairying cows, based largely on data reported in the bulletin noted 

 above. 



The separate inheritance of quantity and quality in cows' milk, J. Wilson 

 (Sd. Proc. Roy. Dublin Soc, n. ser., 12 (1910), No. 35, pp. J,10-Jfl9, dgms. 6).— 

 Using the data of the Ayrshire Cattle Milk Records Committee (E. S. R., 23, 

 p. 780), the author submits diagrams intended to show that there is no correla- 

 tion between quality and quantity in milk. 



Note on the separate inheritance of quantity and quality in cows' milk, 

 K. Pearson {Biometrika, 7 (1910), No. J/, PP- 548-550). — The article noted above 

 is criticized because of the clubbing together of tail frequencies, and because 

 the diagrams failed to show the independence of quality and quantity of millv. 

 The mean percentages and the variation in fat of the 4 groups of Wilson were 

 estimated by the author, with the following results: 



Mean percentctffcs and variations of fat in milk of Ayrshire cattle. 



Quantity of milk. 



Means. 



Standard devi- 

 ations. 



Under 500 gallons 



From 500 to 600 gallons 

 From 600 to 700 gallons 

 Over 700 gallons 



Per cent. 

 3.710±0.010 

 3.673± .007 

 3.651± .008 

 3.640± .008 



Per cent. 



0. 372±0. 007 

 .319± .007 

 .321± .005 

 .275± .006 



" For the whole population of 2,866 cows the mean is 3.672 and the variability 

 is measured by 0.323. It will accordingly be clear that there is (1) a slight 

 but perfectly definite decrease in quality with increase of quantity, the low- 

 yielding cows having a higher, the high-yielding cows a lesser fat percentage 

 than the average Ayrshire, although the increase and decrease only amount to 



