10 OHIO EXPERIMENT STATION: BULLETIN 197 



from three to four weeks. While most of the larvae descended into 

 the sand to pupate, some did not, but transformed wherever they 

 happened to be. Some midges were obtained in each of three glass 

 tubes, containing nothing but cuttings harboring the larvae. The 

 pupa, to casual observance, seemed to differ but little from the 

 larva, except that it had become somewhat shorter. The larvae 

 were seen to disengage themselves from the leaves by their jerking 

 habit and drop to the ground. In the latter part of June the upper 

 layer of earth to the depth of one-fourth to one-half inch was 

 removed from beneath an infested catalpa tree and put into a glass 

 covered case in the laboratory. This earth was kept spread out, 

 the depth being not more than one-half inch, and it was moistened as 

 often as necessary. In early July adult midges appeared in the 

 cage, thus proving that the pupal stage is ordinarily passed in the 

 •ground. These midges were confined on catalpa slips taken from 

 trees grown in the insectary and which had never shown signs of 

 infestation. July 16th, three or four larvae were found on the 

 leaves of one of these slips. No eggs or larvae could be found in 

 the bud of this slip. A different slip, that was put into a jar con- 

 taining midges that hatched from buds on July 4th, had new larvae 

 in the bud July 15th. Some adults, that had issued some time after 

 July 4th, were still living in thfi cage at the time of this observation. 

 The average life of the fly is probably from three or four days to a 

 week or ten days. Sometime in late June both adults and larvae 

 disappeared from the groves outdoors and were not again found 

 until July 16th. In the latter part of summer, the larvae seem 

 chiefly confined to the leaves and seed pods in groves of trees that are 

 of considerable size, but, as before recorded, Mr. Cotton observed 

 them attacking the tender tips of growing nurser}' stock in August 

 and September. Mr. Houser noted, July 30, 1906, that the 

 "terminal borer" was more abundant than ever at Wooster. The 

 dates at which Mr. Houser obtained midges were August 7th and 

 8th and September 3rd. September 12, 1908, larvae were found on 

 leaves and adults were captured from catalpa by sweeping. Prof. 

 Comstock obtained flies about the middle of August from larvae 

 taken in the seed pods. We found larvae in seed pods as early as 

 July 27th, and they still inhabited these pods in mid-September, 

 showing a much longer life cycle than the earlier generations work- 

 ing in the buds. From all the evidence we possess, we infer that 

 there is an indefinite number of generations, the rate of develop- 

 ment probably depending on conditions of heat, moisture and food 

 supply. In the early season the broods are pi-obably distinctly 

 separated, but by the latter part of July all semblance of broods 



