i66 



BACTERIOLOGICAL METHODS 



approximately the same and should occupy a space about 9 mm. in length, being 

 somewhat firmly packed in the lower end of the tube by tapping it sharply on a 

 hard surface. The water in the outer bath should be agitated frequently during the 

 determination. 



Possible Sources of Error. — In applying the foregoing method too great care 

 cannot be exercised with the preparation of the sample. The presence of water, the 

 incomplete solution of the fat in the ether, or the presence of small particles of 

 extraneous matter may interfere with the process of crystallization, frequently caus- 



FiG. 55. — Lard crystals, a, Clus- 

 ters of crystals as seen under the low- 

 power of the compound microscope; 6, 

 crystals highly magnified. 



Fig. 56. — Duck fat crystals, a, 

 Clusters of crystals as seen under the 

 low power of the compound microscope; 

 b, crystals highly magnified. 



ing it to proceed too rapidly and resulting in the formation of a large mass of small 

 fluffy crystals instead of the compact mass of larger crystals desired. These fine 

 crystals render the preliminary washing by decantation with ether difficult, and they 

 also persistently hold the unsaturated glycerids in larger amount than is desirable. 

 The temperature at which the crystallization should be allowed to proceed should not 

 be less than 15° C. nor more than 20° C, with the best results obtainable in the neigh- 

 borhood of an average between the two. Although larger crystals are formed at the 

 higher temperature (20° C), only lards of high grade afford crystalline deposits in 

 working quantity, and in many cases where lards of inferior grades are tested the 

 amount of solid glycerids entering into their composition is so reduced as not to jdeld 

 any deposit at all. 



