Studies in Milk Secretion. 547 



was continued liberall3^ the cows standing in the barn every 

 night. Comparing these records with those of the University 

 herd for a similar period we find no case where the average 

 increase for the herd equals that for the \>rmont herds. In 

 only one year (1893) does it approximate thereto and then only 

 for the two lowest averages found in Vermont. The food con- 

 ditions of the University herd have been given above, where it 

 will be noticed that the barn ration must be fully as nutritious as 

 pasture. Thus far the conditions correspond with those of the 

 Vermont Station. But on turning to pasture the latter kept up 

 the rich barn food while we eliminated nearly all of it, and herein 

 may lie the cause of difference in results. The great increase 

 of the 81 cows is the natural sequence of a change of fresh 

 milkers from poor, dry feed to abundant succulent food, and can- 

 not be justly compared with our conditions. 



As regards quality of milk the Vermont records also show a 

 greater increase in per cent of fat than was found for similar 

 periods at this Station. During practicall}- the same period after 

 going to pasture our cows show nearly the same per cent of fat 

 as when in the barn, while the Vermont herds show an increase 

 of .17 per cent. Whether the differences in food conditions 

 would account for this it is difficult to sa3\ 



Summing up, then, the comparison of records of the two Sta- 

 tions, and throwing out of the consideration the 81 cows, we have 

 three varieties of conditions. First, 105 University cows chang- 

 ing from a rich barn ration to pasture by day and night with 

 most of the former ration discontinued. Second, 39 Vermont 

 cows changing from a rich barn ration to pasture by day with 

 as much of the former ration continued as they will eat during 

 the night. Third, six Vermont cows changing from a relatively 

 poor barn ration to pasture by da}' with the same continuation 

 of the former ration as in the second case. As results we have 

 for the first case, a decrease in quantit}^ and quality of milk after 

 fourteen days of pasture ; for the second and third cases, an 

 increase in the quantity and quality of milk after ten days of 

 pasture. This summary takes for granted that the length of 

 time in milk is practically^ the same for all the cows. In the 

 University herd it is found that some spring cows show no more 

 increase in milk on going to pasture than do some fall cows. 



