898 ANNUAL KEPOKT OF 'I'HJ!] UlT. Doe. 



drawn should, accoi'diiig to quantity taken, be placed upon a clean, 

 smooth Uoor, sheet or paper, thoroughly mixed and then quartered; 

 if the quantity in one of the quarters of the heap of combined sub- 

 samples is too much to submit for analysis, repeat with it the pro- 

 cess to which the larger (luantity of material has been subjected; 

 and so on, until the quantity is reduced to a convenient amount for 

 mail or express. Of fine-ground materials, four to eight ounces is 

 usually sufficient; of coarser feeds, send double or treble this amount. 



There are many cases, however, in which the quantity sold of any 

 particular mixture is so small that the expense of analysis seems 

 relatively high; and some other method of fixing upon a safe guar- 

 anty is sought. The composition of the mixture can readily be com- 

 puted from the weights of the several ingredients and their re- 

 spective percentage compositions, where these facts are known. 



The ingredients commonly employed may be regarded as belong- 

 ing to two quite distinct classes, the first, including whole grains or 

 their milling products prepared under the ordinary conditions of 

 milling by the mixer himself; the second, including x>roprietary mix- 

 tures or other materials bought from parties and concerning whose 

 composition often little is certainly known. 



Jenkins and Winton gathered together some years ago, the re- 

 sults of American analysis of the first class of ingredients, and pre- 

 sented them in tabular form so as not only to. show their average 

 composition in the more important constituents, but also the ex- 

 treme variations in percentage of the several constituents that have 

 been found in America. An abstract of that table is appended to 

 this circular. (See Table No. I). 



No such exact knowledge is possessed concerning the second class 

 of ingredients, not only because they have been produced by others 

 than the mixer and are therefore subject to adulterations of which 

 he can have no knowledge, but also because the materials are often 

 mixtures that vary in their composition because of a change in the 

 manufacturing methods as a result of which they are produced; or 

 . that are varied by the makers from time to time as their profit may 

 indicate. The local mixer is, in such case, compelled either to buy 

 them under guaranty of composition from responsible parties or to 

 have them analyzed. The results of analysis of a large number of 

 such of these feeds as find their way to the retail market are pre- 

 sented in Bulletin No. 81 of the Department, which can be had upon 

 application. 



HOW TO USE THE TABLES OF ANALYSIS. 



1. To deter mine lupon a percentage for guaranty: All the raw oi- 

 unmanufactured feeding-stuffs are seen, by reference to the tables, 

 to vary considerably in composition; there is no certain iiidlcalion 



