1898.] PUBLIC DOCUMENT — No. 33. 77 



alluded to, has been reported to them. While more work 

 will be done along this line, it has been quite clearly demon- 

 strated that the so-called Maercker and Reinke methods for 

 the estimation of starch in agricultural products are faulty, 

 and will give altogether too high results. The only method 

 from which reasonably accurate results may be expected is 

 the diastase or malt method,* and this method has been 

 adopted by the official chemists in place of all others. 



Part II. 



Experiments with Pigs. 

 Two experiments have been completed with pigs, and a 

 third is now near completion. These experiments were 

 designed to study the value of corn meal as compared with 

 hominy and cerealine feeds for pork production, when fed in 

 combination with skim-milk. Both these feeds are quite 

 similar in composition. They consist of the hull, germ and 

 more or less bran and starch removed from white corn, dur- 

 ing the preparation of cracked hominy and cerealine flakes 

 for human consumption. Cerealine is much more bulky 

 than the hominy feed. These experiments will be published 

 in detail later. It can be said, however, that pigs have made 

 nearly, and in some cases fully, as good growth on these 

 feeds as on an equal amount of corn meal. 



Salt Marsh Hay. 



A thorough investigation has been completed concerning 

 the general character and feeding value of salt marsh hay. 

 The results are being published in bulletin form. The prac- 

 tical conclusions, briefly stated, are as follows : — 



The several varieties of salt hay have, ton for ton, from 10 

 to 17 per cent, less feeding value than average English hay. 

 When 10 to 12 pounds of salt hay were fed daily, together 

 with 7 or 8 pounds of grain and a bushel of ensilage, the 

 ration produced within 2 to 5 per cent, as much milk and 



* Sachsse's method can be used for estimating starch in commercial starch and in 

 potatoes. 



