REPORT OF THE CHEMIST, 



57 



send their products to market until the water has been reduced to 12 

 per cent, or less. 



CUED. 



I have tried many ways to estimate the curd. To filter the butter 

 through paper washed with ether or naptha, and wash the curd into a 

 tarred dish, dry and weigh, ignite and deduct weight of ash, I have 

 found unreliable. 



To filter through a Gooch crucible aud proceed as above (without 

 washing) is better, but a very slow process. Moist fat, even dissolved 

 in ether, filters with difficulty through a Gooch. To fill the crucible 

 partly with sand helps the process, and is a moderately good method. 



Fair results are obtained by using a tarred filter and weighing the cu i d, 

 after drying, on the filter. After ignition, the weight of the salt is to be 

 deducted, and percentage of curd determined. 



Better still, and the best method found^is to dry in porcelain or 

 platinum dish without sand. The dish should have a small stirring 

 rod, be heated for two hours at 105°, stirred every twenty minutes. If 

 drops of water are still seen on bottom of dish, a little absolute alcohol 

 is to be stirred in and the dish reheated. Ether or naphtha is then to 

 be added, and the solution filtered through a Gooch crucible. All tho 

 curd is to be carefully washed into the crucible with an ether wash 

 bottle, crucible dried for an hour at 105°, and weighed. This gives 

 total weight of salt and curd. From this deduct the weight of salt, 

 determined as hereafter described. Eemainder equals weight of curd. 



CASEIN. 



Curd is composed essentially of casein. I have thought, however, 

 that if the real amount of albuminous matter present could be deter- 

 mined it would be more useful than to know the total amount of curd. 

 With this purpose in view, there was made a series of determinations 

 by combustion with soda-lime and by moist combustion with alkaline 

 solution of permanganate of potassium. 



The numbers obtained are in the following columns: 



Xumber. 



1 



2 

 3 

 4 

 5 

 C 



From these results it is seen that the two methods, while not giving 

 identical results, nevertheless present an agreement as close as could 

 be expected from the character of the manipulations. 



In the soda-lime process about five grammes of the butter were taken. 

 The amount of gas evolved during the combustion was very consider- 



