286 Journal of Agricultural Research voi.xvn.No. 6 



anything approaching a sound basis for such a judgment is a thorough 

 analytical study, with the best of biometric tools, of the normal varia- 

 bility of milk and fat production. 



MATERIAL FOR INVESTIGATION 



The present study is based on the records of Ayshire cattle published 

 in the Reports of the Ayrshire Cattle Milk Records Committee of Scotland, 

 compiled by Speir {26) ^ and Howie ^ (6), Portions of the very valuable 

 records gathered by this committee have been used by other students of 

 the problems of milk production, notably Wilson {30), Pearson {23), and 

 most recently Vigor {28). Wilson made use of the 1908 records, and 

 Vigor those of 1909 for the Fenwick district only. 



The reports under consideration include, so far as it is possible to get 

 the information, the following items : 



1. Total milk produced (in gallons). 



2. Average percentage of fat, determined from periodic tests. 



3. Total milk calculated to a 3 per cent fat basis. 



4. Weeks in milk. 



5. Age of cow. 



6. Date of last calving. 



7. Miscellaneous information about the cow, particularly of abnormal 

 circumstances of any sort during the test. 



In many cases information is lacking on some one or more of these 

 points, so that, while altogether 8,132 cows were tested in 1908 and 9,202 

 in 1909, nothing like these numbers are available for analytical study. 

 Another difficulty arises in the fact that there is, of course, much over- 

 lapping of calendar years by the lactation periods. Again there is in 

 some districts frequent failure to state the age of the cow. 



In the present study all available records from the 1908 and 1909 

 reports have been used, if they came within the following regulations 

 which we established in order to secure critical material for variation 

 study: 



(a) The recoid must be complete in all particulars — that is, cover 

 items I to 6 in the list above. 



(6) The record must be based on 32 or more weeks in milk. 



(c) There must be nothing of an abnormal or unusual nature about 

 the cow or the lactation, so far as discoverable from the records. 



The first of these restrictions requires no comment. 



Regarding the second it may be said that the reason for imposing 

 this restriction was that, for present purposes, we desired to use long 

 term averages, rather than to consider lactations of all durations. There 



' Reference is made by ntunber (italic) to "Literature cited," p. 320-322. 



' It is a great pleasure to acknowledge, with grateful thanks, the kindness of Mr. John Howie, of Ayr, 

 Scotland, the secretary of the Milk Records Committee, in furnishing a set of the committee's reports for 

 this investigation. 



