MILK PRODUCTION 477 



milk production, and render the interpretation of their results 

 peculiarly difficult. It is obvious, e.g., that if a change of the 

 ration of a cow is accompanied by a decrease in her milk yield 

 part at least of the decrease may be due to the- progress of lac- 

 tation and not to the change of feed. On the other hand an 

 increase of the milk yield in a later period of the experiment 

 may be partly offset by the natural shrinkage in milk. In 

 brief the later periods of an experiment are at a disadvantage 

 compared with the earlier periods. 



Two methods for eliminating or attempting to eliminate this 

 influence of lactation have been used, viz., the period system 

 and the group system. 



571. The period system. In the period system, as intro- 

 duced by Wolff, Klihn and others of the earlier experimenters, 

 the animal receives an identical ration in two or more periods 

 well removed from each other in point of time usually the 

 first and last periods and from the results of these periods the 

 average daily rate of decrease in the yield of milk and its in- 

 gredients is calculated. On the assumption that had the same 

 ration or treatment been continued unchanged this rate would 

 have been uniform throughout the experiment, it may be 

 computed what yields would have been secured in the inter- 

 mediate periods. A comparison of these computed yields with 

 those actually observed is taken as the measure of the effect 

 of the change in feed or other conditions. The accuracy of 

 this method depends of course on the correctness of the as- 

 sumption that the yields would have decreased at a uniform 

 rate. 



572. The group system. The use of the group system was 

 introduced by Fjord and his successors in the Copenhagen 

 Experiment Station in connection with their determinations of 

 the so-called feed units (702). The period system seeks to 

 compare each animal with itself. The group system, on the 

 other hand, attempts to compare an animal or group with an- 

 other check animal or group. In a long preliminary period 

 both groups receive the same ration or treatment and their 

 relative production is determined. One of the groups is then 

 continued on the same treatment while with the other group 

 the factor to be tested is introduced. Finally, in a concluding 

 period, both groups are again treated as in the initial period. 



