214 INFECTION AND IMMUNITY. 



TO obtain To obtain serum for the test one may resort to 

 $emm. j^g^^g . pj ace a ca ntharides plaster from one- 

 half to three-fourths of an inch square on the ab- 

 dominal skin, protect it with a dressing, and in 

 about twelve hours remove the serum with a steril- 

 ized hypodermic syringe. Or, one may collect in a 

 small test tube from 0.5 to 1 c.c. of blood from the 

 lobe of the ear or finger-tip, and draw off the serum 

 after it has separated by clotting. It is the cus- 

 tom in some well-equipped laboratories to fill sev- 

 eral U-shaped capillary tubes with blood from the 

 lobe of the ear and to separate the blood from the 

 serum at once by centrifugation. The custom of 

 drying a few drops of blood on a coverglass or on 

 filter paper, and of sending this preparation to a 

 laboratory for the agglutination reaction, has been 

 practiced quite extensively, and is a justifiable 

 procedure when it is not possible to collect the pure 

 serum. It has the disadvantage that the experi- 

 menter never knows exactly how much blood has 

 been collected, and consequently can not perform 

 the test with exact dilutions of the serum, the im- 

 portance of which will be pointed out below. The 

 red corpuscles and debris in such a preparation also 

 interfere with the clearness of the field in micro- 

 scopic examination, a difficulty which may be 

 partly overcome by filtering the dissolved serum. 

 semm When only a small amount of serum is available, 



Dilutions, y. . g necessar y to use the microscopic method. 

 Normal human serum, when concentrated, and 

 even when diluted to one in ten or higher, some- 

 times agglutinates the typhoid bacillus and some 

 other organisms; the same serum, when diluted to 

 one in forty or one in sixty, may not agglutinate. 

 The serum of a typhoid patient, however, or of a 



