BIOLOGY OF LIMNEKIUM VALIDUM. 91 



rate the series into the two species, so that some doubt remains 

 whether L. pilosulum was actually represented. 



The two most important lots of the larvae were obtained respec- 

 tively on August 18, from third-stage hosts collected the day pre- 

 viously at Eeading Highlands, Mass., and on September 6 to 9, from 

 mostly sixth-stage hosts collected near the laboratory at Melrose 

 Highlands. The former lot may have been L. pilosuluTJi, as this 

 species spins its cocoon inside of the skin of small hosts and issues 

 therefrom throughout the month of September. The latter lot of 

 larvae seems to have been L. validuTn without any doubt, as they were 

 in the first stage when L. pilosulum was already beginning to issue 

 as imagos, and inasmuch as they were found in rather large sixth- 

 stage hosts. It seems reasonable to conclude, therefore, that the first- 

 stage larvae of these two species are practically identical in structure 

 and appearance. 



The disposition of the larvae in their natural host is not without 

 interest. According to Mr. Tothill's notes, they were found free 

 in the body cavity, between the walls of the body and the alimentary 

 canal, either inferior or superior to the latter. No indication of 

 amcebocytosis was encountered during the course of the dissections. 



GENERAL SUMMARY. 



Limnenu'nh iml'idum. (Cresson), a common parasite of the fall web- 

 worm {Hyphantria cvrwa Drury), readily attacks the caterpillars 

 of the brown-tail moth, gipsy moth, and rusty vaporer moths, and 

 also the tent caterpillar, when placed in confinement with these 

 hosts, but is able to complete its transformations in the last species 

 only and even then in but a small percentage of cases. Its larvae 

 seem to be totally unadapted for life in the caterpillars of the three 

 former species, and fail to survive the protective reactions of the 

 liost, wdiich are visibly manifested by an accumulation of active 

 blood cells or amcebocytes around the larvae, the cast eggshells, and 

 even the eggs themselves. The amcebocytes presumably attack tlie 

 living eggs and larvae, or at least ultimately efface the latter entirely. 

 The same reaction takes place in the case of the tent caterpillar, 

 but a few of the larvae are able to complete their transformations. 

 Adaptation here is partially in evidence, and may be due to larval 

 secretions which ward off the protective reactions of the host. 



The Qgg of Lirmierium has a thick, chitinized chorion, which re- 

 sists decomposition in the body fluids of the host to a remarkable 

 degree. The first-stage larva is characterized by a long, tapering 

 tail appendage, evidently adapted for a respiratory function and 

 acting as a blood gill. The second-stage larva has a much shorter 

 appendage, and the third-stage larva has none, so that respiration 

 very likely occurs through the walls of the digestive tube, oxygen 



