May IS, 191S Wilt of Gipsy-Moth Caterpillars 115 



It will be seen from Table I that the blood, in general, is a fairly reliable 

 index of a caterpillar's condition. Still, it must be noticed that under 

 tests 9, 10, and 23 the conditions in the blood did not represent those of 

 the tissues. However, not a single case was found where polyhedra were 

 in the tissues, but not in the blood. Tests 9, 10, and 23 also indicate that 

 the blood is the first tissue to be affected. This is to be expected, since 

 infection, as will be shown later, occurs by way of the alimentary tract. 

 The virus is ingested with the food and possibly passes through the 

 epithelium of the intestine into the haemolymph, by means of which it is 

 distributed to the other tissues. 



In 1 91 3 both Dr. Chapman and the writer were fully aware of the 

 impracticability of the blood test and also did not consider it an accurate 

 means of diagnosing diseases of caterpillars. After more extensive studies, 

 however, the conclusion has been reached that the blood test, though not 

 irreproachable, is, nevertheless, a better index than it seemed at first. 

 If used at all, the test must be made with extreme care, but, owing to the 

 length of time it requires, it can not be used when one is working ^\dth 

 several hundred caterpillars. 



ETIOLOGICAL INVESTIGATIONS 



In 1 91 3 the problem was again attacked in the most critical manner 

 from the bacteriological point of view. An account of all the work under- 

 taken will not be given, as the results were entirely negative, not essen- 

 tially differing from those obtained by Jones (10). After a caterpillar has 

 been dead for some time, bacterial invaders enter and cultures are not diffi- 

 cult to obtain; but the writer found, as Jones had, that, when attempts 

 were made to isolate bacteria from caterpillars that had recently died in 

 the field, the tubes remained sterile in most cases, and that, v/hen a growth 

 was obtained, the different species of bacteria isolated behaved as a non- 

 pathogenic intestinal flora. Among all these attempts, no growth was 

 obtained in tubes inoculated with material taken from the body cavity of 

 Hving diseased animals. The media used, at varying degrees of acidity 

 and alkalinity, included nutrient veal and beef agar and gelatin, bouillon, 

 potato, and caterpillar soup. Anaerobic cultures were also tried, but the 

 results w^ere likewise negative. The bacteriological results bore a striking 

 similarity to those of Tangl (13) and Wachtl and Komauth (17) with 

 Wipjelkrankheit and to the results of Jones (10) in his work on the 

 gipsy moth. 



After having satisfactorily eliminated the higher parasites (nematodes, 

 protozoa, etc.) and after obtaining negative results bacteriologically, the 

 following possibiUties suggested themselves: (i) Either the polyhedral 

 bodies themselves are parasites; or (2) the etiological factor of the wilt 

 disease is a minute filterable organism independent of the polyhedra; or 

 (3), if the virus is filterable, it may be genetically related to the polyhedra. 

 The filterable-virus viewpoint seemed very promising, especially since 



