872 EXPERIMENT STATION" RECORD. 



of milk," and of assuming 30 daily percentages of butter from only two, it is 

 concluded that " the trustworthiness of the advanced register records is very 

 slightly impaired by the use of only a two-successive-days' test to sample the 

 entire 30 of a month. It may be assumed with practical certainty that such 

 sampling will never cause an error of over 23 lbs. in the total result for a 

 year, that in 99 cows out of a 100 the error will be under 15 lbs., and that the 

 average error will be 4^ lbs. under. The divergence of a cow's record, 12 one- 

 day tests being used, from her true record, will be under 6i lbs. on the average, 

 under 21 lbs. in 99 cows out of a 100, and under 32 lbs. always." 



From a study made of the records of 50 of the cows with respect to the pos- 

 sible divergence during a term of years, "it appears that a year's record as 

 now taken by 12 two-day tests would never d'verge from the average of 15 

 years' records similarly taken for that cow (were such obtainable) by more 

 than 35 lbs., would diverge therefrom by more than 22 lbs. or more only once 

 in a hundred times, and would diverge therefrom on the average by a little 

 under 7 lbs." This 35 lbs. may then be taken as the upper limit of possible 

 error since it also includes the divergence that might occur during the single 

 year. " It seems safe to credit at least 7 of the 35 lbs. to these sources of 

 variation which are added to those which make a record as now determined 

 vary from the cow's true record for the year. If this is done we have then an 

 upper limit of error for the record from 12 two-day tests of 28 lbs. butter fat 

 by this method. ... To say that in 99 cases out of 100 it will be true within 

 18 lbs., and that on the average it vsill be true within 5^ lbs., is to make a 

 thoroughly guarded statement." 



Comparative value of one- and two-day tests as viewed by the experiment 

 station {Guernsey Breeders' Jour., n. ser., 4 (1913), No. 3, pp. 16, 17). — In 

 connection with the endeavor of the Jersey and Guernsey Associations to get 

 a uniform basis of advanced register testing, inquiries were sent to 24 experi- 

 ment stations asking the comparative cost and relative accuracy of one- and 

 two-day tests. The answers indicated that the difference in cost is slight and 

 the accuracy of the two-day test greater. 



The nature and causes of the variations in the richness in butter fat of 

 milk from the same cow, E. L. Tiiorndike (Guermcy Breeders' Jour., n. ser., 

 4 {1913), No. 5, pp. 35-39, figs. 6). — In connection with the milk fat tests made 

 by the American Guernsey Cattle Club, reported above, a study was made of 

 the causes of the variations in richness of the milk fat. It was found that a 

 cow has a general average richness of milk which distinguishes her from other 

 cows. If this general average of richness is taken as 100, the climatic, feeding, 

 and other conditions related to season of the year, make her tend in any 

 month to vary from this general average richness as follows: January 104.9, 

 February 103.1, March 100, April 99.4, May 98.6, June 95.5, July 95.7, August 

 97.4, September 98.8, October 101.2, November 101.3, and December 104.5. 

 These figures relate to the northern states from which tests were taken. The 

 stage in the lactation period makes her tend to vary from this general average 

 richness as follows: First month of lactation 89.6, second month 90.3, third 

 92.4, fourth 96.5, fifth 97.9, sixth 101.2, seventh 103.2, eighth 103, ninth 104.3, 

 tenth 104.9, eleventh 105.3, and twelfth 109.4. However, this general tendency 

 which the cow has as a member of the breed and species seems to be much 

 complicated by individual peculiarities. 



Of the cow and her treatment, " some cows seem to give almost as rich milk 

 at the start as at the end ; some to change gradually all along the year ; some 

 to change very rapidly in the early months and then hardly any ; some to change 

 hardly any till the later months and then very rapidly. There are in addition 

 to these long swings and changes, irregular temporary ups and downs due to 



