126 Journal of Agricultural Research voi. iv. No. 2 



could constantly soil the food and infect the living ones. Figure 1 7 gives 

 the mortality among these insects. Twenty-one caterpillars died of the 

 wilt disease, ten of the "other cause," eleven of tachinid parasitism, and 

 five male and three female moths emerged. The experiment covered 

 a period of 32 days, an ample length of time for every individual in the 

 tray to become infected; yet in spite of the 21 deaths from wilt, 8 

 moths were obtained. 



SUMMARY 



(i) The wilt of gipsy-moth caterpillars is a true infectious disease that 

 is distributed over the entire territory infested by the gipsy moth. 



(2) Epidemics of the disease occur only in localities heavily infested 

 by the gipsy moth. 



(3) Climatic conditions appear to bear an important relation to wilt 

 in the field. 



(4) The disease is more prevalent among older than among younger 

 caterpillars, but small caterpillars also die of it in the field. 



(5) No diagnosis of wilt is valid unless polyhedra are demonstrated 

 microscopically. 



(6) There is no account of the occurrence of wilt in America prior to 

 1900. 



(7) Minute dancing granules may be observed in wet smears. 



(8) Polyhedra are probably reaction bodies belonging to the highly 

 differentiated albumins, the nucleoproteids. 



(9) The pathology of wilt does not vary with the age of the caterpillars. 



(10) The polyhedra originate in the nuclei of the tracheal matrix, 

 hypodermal, fat, and blood cells. 



(11) The nuclei of the tracheal matrix and blood cells seem to be the 

 first tissue nuclei affected. 



(12) Many minute violently dancing granules are found in the patho- 

 logical nuclei of fresh tissue. 



(13) Giemsa's stain demonstrates many little granules in the nuclei of 

 diseased tissue sections. 



(14) The alimentary canal seems to be the last organ in the body to 

 disintegrate. 



(15) Two types of blood corpuscles exist in normal haemolymph. 



(16) Two types of pathological blood corpuscles exist in diseased 

 caterpillars. 



(17) The blood is a fairly reliable index of a caterpillar's condition. 



(18) The blood test is impracticable for large experimental series. 



(19) Bacteria are not etiologically related to wilt. 



(20) The virus of wilt is filterable with difficulty. 



(21) Such a filtrate is free from bacteria and polyhedral bodies. 



(22) Caterpillars that have died from infection with filtered virus are 

 flaccid, completely disintegrated, and full of polyhedra. 



