TESTS AT EXPERIMENT STATIONS. 



119 



phosphoric acid, equal to 70 per cent of the market value of the feed fed the 

 animals." Of course, even by best methods, there would be some waste of 

 manure. Estimating that waste at the largest reasonable quantity, it still 

 remains true that the total omission of this item is extremely unjust to the 

 heavier feeding animals. On page 40 of the seventh annual report of the Ver- 

 ment Experiment Station we find an estimate of the combined feeding and 

 fertilizing value of skim-milk of thirty-six cents per 100 pounds. This is 

 approximately four cents per pound for solids not fat, probably a higher valua- 

 tion than can be realized in ordinary practice. Whatever may be deemed a fail- 

 valuation, its inclusion is absolutely necessary in a test of dairy cows. The 

 approximate valuation of skim-milk, of manure and of butter is as easy as 

 that of the foods consumed, and is equally necessary to the conduct of a breed 

 test. 



The breed test of the New Jersey Experiment Station was brought to an 

 unhappy end November 1, 1890. Nevertheless, its conductors considered that 

 it afforded ample ground for wide generalizations. " It is assumed," say they, 

 page 32, Bulletin 32, " that the individual animals do represent their respective 

 breeds, and that the average daily results secured from the herds of animals 

 selected as representatives of their breeds are relatively the same as might be 

 expected from a representative animal of each." Unless this assumption is 

 justified, the conclusions are worthless. On this point I mention just two 

 facts. Out of the five breeds represented Ayrshire, Guernsey, Holstein- 

 Friesian. Jersey and Shorthorn the smallest quantity of milk was produced 

 by the Jerseys, and next to the smallest (about fifteen per cent less than the 

 next larger) was produced by the Holstein-Friesian. The milk record of one 

 of the Holstein-Friesian cows for the period May 1, 1889, to October 31, 1890, 

 was 3,824 Ibs. an average of about 7 Ibs. per day for this "representative" 

 Holstein cow. Upon such a foundation are built the generalizations regarding 

 entire breeds of cows. If the engineers of the coast survey were to guess at 

 the length of their base line, their subsequent work would possess about the 

 same scientific value as that in question. 



The only other specific breed test of which I have any report is that of the 

 Maine Station. The number of the animals and breeds is smaller than at either 

 the New York or New Jersey tests. The methods are open to some of the same 

 objections. The premises are altogether too narrow to support the generaliza- 

 tions. For the purpose, however, of illustrating what, in my judgment, a 

 report of such an experiment should show, I append a table giving the results 

 of this station for one year. The table is not copied, but is a compilation from 

 the tables contained in the Maine report for 1890. 



Note H.-F., Holstein-Friesian; A., Ayrshire; J., Jersey. 



Butter is assumed to be worth twenty-five cents per pound. The quantity is 

 computed by adding one-sixth to the quantity of butter fat. Milk solids not fat 

 are assumed to be worth 2i cents a pound. It is also assumed that forty per 

 cent of the value of the food consumed is saved in the manure. As to the 

 fairness of this last assumption, see an article by Dr. Collier in the Country 

 Gentleman for December 20 last, page 916, in which he largely quotes Dr. 

 Goessman's bulletins from the Massachusetts Station. 



The age, weight and date of calving and average per cent of fat should 

 also be included, but were not given in the 1890 report. Full details as to feed- 

 ing should also be given. The Maine experiment was continued another year 



