139 



of 28 fly larvae, approximately 193 pupae and approximately 

 196 adults examined, only two larvae were found to contain 

 one worm embryo each (experiments 7 and 8), and one adult 

 one malformed larva (experiment 14). In further experiments 

 (Nos. 15 and 16) larvae of H. domestica were allowed to 

 develop in sterilized faeces to which larvae of both H. muscae 

 and H. microstoma had been added. 



In experiment 15, of 28 flies examined 15 were found 

 infested with worm larvae. In Table 5 particulars are given of 

 6 larvae obtained from these flies. Hill believes that 4 of 

 these larvae, specimens 3-6, are larvae of H. microstoma. The 

 evidence in favour of these larvae being those of H. microstoma 

 is not entirely convincing, particularly as the final larval 

 stage was not observed, and the present writer doubts the 

 correctness of Hill's conclusion. 



Nevertheless, one does not doubt the possibility of M. 

 domestica acting as an intermediary host of M . microstoma, 

 but more evidence is required before proof of such is estab- 

 lished. 



In commenting on the present writer's preliminary 

 observations, Hill (p. 62) casts doubts upon the results 

 obtained in the experiments with larvae of H. megastoma. In 

 these preliminary observations the opinion was expressed that 

 it would be difficult or impossible to differentiate with absolute 

 certainty between the final larval stage of II. muscae and that 

 of II. megastoma. 



Hill, unfortunately, failed to appreciate the fact that this 

 opinion was expressed from the point of view of the possibility 

 of differentiating larvae obtained from habronemic granu- 

 lomata, and he further missed altogether the reference to the 

 fact that differences had been observed, particularly with 

 regard to the length of the oesophagus. His inference, there- 

 fore, that the present writer was not dealing with pure cultures 

 cannot be held to be correct. 



The fact that Hill (p. 64) failed to satisfy himself as to 

 the specific determination of larvae obtained from conjunctival 

 lesions, but considered they resembled those of H. megastoma 

 more closely than those of II. muscae, seems to support the 

 conclusions outlined in the present communication, viz., that 

 it would be difficult or impossible to differentiate between the 

 final larval stage of II. muscae and H . megastoma, except 

 under the best conditions of preservation, etc., and that evi- 

 dence is in favour of the probability that larvae of H . 

 megastoma are more often responsible for the production of 

 habronemic granulomata than the larvae of the other two 

 species. 



