220 Bulletin 397 



basis of probable productive capacity- From the discussion of the factors 

 causing variation in milk production, the great uncertainty involved in 

 selecting groups of equal production is evident. In this connection Hills 

 (1907) states: " He who can, from an often limited number of animals, 

 formulate groups which for a great length of time will prove essentially 

 equivalent in their milk-making powers is gifted with second sight." 

 This is a very serious disadvantage of the continuous system. However, 

 this difficulty should be minimized by the application of statistical 

 methods. The variable factors due to individuality cannot be eliminated, 

 but the probable limits of variation can be ascertained. 



CONSIDERATIONS IN GROUPING EXPERIMENTAL ANIMALS 



In selecting two groups of animals for use in an experiment conducted 

 under the continuous system, it is clearh^ recognized that the two groups 

 should be as nearly alike as possible. Points that must be considered 

 are: previous treatment and history, age, stage of lactation, breed and 

 type, productive capacity. 



In this connection it is believed that a consideration to which insuffi- 

 cient attention has been given is that of history as regards diseases. 

 A diseased animal is not nonnal and should not be included in a feeding 

 experiment. It goes without saying that an animal used in experimental 

 work should pass the tuberculin test. The herd records should furnish 

 data as to abortion and breeding troubles, in order that animals abnormal 

 in these respects may be excluded. Finally, it is believed that each indi- 

 vidual selected for experiment should be given an examination by a 

 veterinarian as to general health. 



Considerable study has been made of the experimental work that has 

 been done in regard to productive capacity, in an attempt to ascertain 

 the points considered in selecting animals on this basis. The majority 

 of investigators, however, have made little mention of the matter, and 

 most frequently one finds merely the statement that " groups of equal 

 production were chosen." Usually milk production is meant. Occa- 

 sionally one finds the statement that groups were made equal as regards 

 both milk and fat production. Just how this would be possible within 

 very close limits with only a moderate number of animals to select from 

 is not clear. Another ambiguous point is, just what is meant by " equal 

 production." No statement has been found of the limits that have been 

 allowed in choosing equal groups on the basis of production records. 



Some, investigators have selected groups on the basis of yield during 

 the previous lactation period. That this alone is a poor criterion is shown 

 by the work of Gavin (1913b). He found that the estimation of one 



330 



