4 PROCEEDINGS OF THE ACADEMY OF [Jan., 



in each and every case the Coccidia happened to be in precisely such 

 a stage as to resemble the Sarcosporidia. The mathematical proba- 

 bility of this taking place diminishes with the number of mice used 

 and when this number is large becomes a vanishing quantity. 



Of course, since protozoan infections naturally tend to assume the 

 epizootic form, if one or two mice from a given cage were found to be 

 infected with Coccidia, the surmise would be warranted that many 

 or all of the others were so infected. But in the case of the series 

 upon which the present studies were based, the mice were obtained 

 from various sources and from various places and in many cases had 

 never been in contact. It is therefore believed that the results as 

 stated are valid, so far as concerns the possibility of confusion with 

 Coccidia. 



The mice selected for inoculation were deprived of food for 24 

 hours. This served a two-fold purpose. It rendered them more 

 prone to eat the infectious material when given and it served to free 

 the intestine of half-digested vegetable food, the presence of which 

 interferes with sectioning. For the short periods, up to 18 hours 

 or so, the mice were given a piece of infected muscle of what was 

 regarded as an appropriate size. This was larger or smaller, according 

 to the number of cysts it contained; and when these were abundant 

 the portion given had a weight of the order of one-tenth of a gram. 



When the mouse is given its infecting meal, any one of several 

 things may happen. Some mice positively refuse to touch the 

 meat, while others merely play with it for a time and then abandon 

 it. More usually, however, the mous£ feeds, and the customary 

 method is for it to hold the food in its forepaws and nibble at it 

 until it is all consumed. This procedure, however, may be inter- 

 rupted by delays, but if the entire time required to finish the meal 

 is short in comparison with that to elapse between feeding and 

 death, the mouse may be used. Finally, in some cases the meal was 

 bolted in the manner in which a dog feeds. 



At the end of the proper period the mouse was chloroformed, 

 opened, and the alimentary canal removed. In nearly all cases it 

 was placed in the killing fluid entire, being neither cut into pieces 

 nor slit open. The intestine of a mouse has such thin walls that the 

 fluids have no difficulty in penetrating, and this procedure does away 

 with the rough handling necessary in slitting the intestine. It more- 

 over retained the intestinal contents, an obvious advantage, and a 

 comparison with slit intestines showed that the fixation was equally 

 accurate. The only disadvantage was that at times the penetration- 



