1918] Coccobacilliis Acridiorum D'Herelle 31 



still it seems to me that the organism acts very quickly. These 

 "per os" infections really mean more than the inoculation 

 experiments for the reason that it is the natural way in which 

 the bacterium would invade the insect. Of course, laboratory 

 passages, where pure recoveries are required, are impossible to 

 perform by this method of infection unless one plated between 

 each infection. Since my experiments showed the futility of 

 passages, in so far as increase in virulence is concerned, I did 

 not see any advantage in doing an extra amount of tedious 

 work. It seems to me that the organism is sufficiently virulent 

 even in old cultures, so that if one could succeed in establishing 

 a center of infection in the field an epidemic would soon follow 

 provided certain conditions were favorable. 



Tables XX and XXI also represent food infections. M. 

 bivittatiis was the subject. If it is permissible to judge from 

 two experiments the organism does not seem to be so patho- 

 genic to this insect. A number of insects in the XX experiment 

 succumbed to parasitism by Mermis ferruginea, a nematode. 



Table XXII represents the same sort of an experiment as 

 the preceding with the exception that M. femur-rubriim was 

 the subject. "Souche Cham" also does not seem to be as highly 

 pathogenic to this species as it is to M. atlanis. 



Table XXIII represents an inoculation experiment with 

 "Souche Sidi." M. atlanis was the subject. The period from 

 infection to death extends over a period of six days. This 

 seems to show that "Souche Sidi" is not as pathogenic as 

 "Souche Cham." 



Table XXIV represents a food infection experiment with 

 the same culture and the same subject. The period from 

 infection to death is also, in general, prolonged. Two animals 

 died naturally although I am certain they became infected. 



In all of the food infection experiments the grasshoppers 

 were given barely enough leaves in order to insure their eating 

 everything in 12-24 hours. 



Table XXV represents a food infection experiment with 

 "Souche Sidi." The subject in this case was M. bivittatus. 

 The pathogenicity of "Souche Sidi" for this species seems to be 

 the same as for M. atlanis. 



