210 THE AMERICAN MONTHLY [June 



blue, and even with these the bacilli stain more deeply at 

 the ends than in the middle, so that they appear some- 

 thing like diplococci. 



For the demonstration of the bacilli in the blood, 

 Canon recommends a rather complicated method. The 

 blood is spread upon clean cover glasses in the usual way, 

 thoroughly dried and then fixed by immersion in abso- 

 lute alcohol for live minutes. The stain which seems best 

 is Czenzynke. 



R Concentrated solution metliylene blue 40 parts. 



0.5 per cent solution eosine in 70 per cent 



alcohol 20 " 



Distilled water 40 " 



The cover glasses are immersed in this solution and 

 kept in the incubator from three to six hours, after which 

 they are washed in water, dried and then mounted in 

 Canada balsam. 



By this method the erythrocytes are stained red, the 

 leucocytes blue, and the bacillus, which is also blue, ap- 

 pears as a short rod, or even as a dumb-bell. The bacil- 

 lus does not grow in gelatine or upon ordinary agar. 



We encounter quite a difficulty to describe a certain 

 set of symptoms, for the type of the attack varies from 

 time to time in different localities, so that we rely in the 

 diagnosis of this disease on various factors, chiefly the 

 one that influenza is epidemic at the time, and perhaps 

 that other members of the household have sufl"ered. The 

 diagnosis must therefore be made by a process of exclus- 

 ion in very young infants. 



We need not be surprised to find various types of this 

 disease in infancy ; as previously mentioned in this 

 paper, namely, the form known as gastro-enteric type and 

 the pulmonary type, for we find that the ordinary so- 

 called pneumonia diplococcus can and frequently does 

 cause at one time an otitis, at other times a meningitis. 



It is in this manner that the influenza bacillus some- 



