240 SCIENTIFIC FEEDING OF ANIMALS 



therefore, not free from objections, and the results 

 are more reliable when the trials are carried out 

 upon groups of animals at the same time. Here 

 also the chief condition is that the groups which are 

 to be compared with one another should be equiva- 

 lent at the beginning of the investigation as regards 

 race, age, sex, live weight, etc., of the animals, in 

 fact, one group must be the counterpart of the 

 other. When such groups have been arranged, 

 a preliminary trial must be made to see if they 

 behave similarly on the same food. Should differ- 

 ences be observed changes must be made in the 

 groups until a close agreement is found, afterwards 

 the food to be investigated may be given. 



Investigations, then, are of two kinds, those 

 conducted in periods and those in groups ; the 

 former have only a limited use, for the animals 

 change in condition, especially young and fattening 

 cattle, and so the conclusions are not reliable. 

 With milking cows, though, this method can be 

 used, and mention will be made of it later. 



With regard to the kind and quantity of the food 

 used in the trial it is impossible to give any general 

 rules, for each investigation has different objects. 

 Where a feeding-stuff is being tested the total 

 quantity of food must not be too large, because 

 with excess the nutrients in different rations would 

 not show any difference at all. If the object is to 

 see how varying quantities of protein in the food 



