THE DOCK FALSE-WORM. 11 



satisfactory, since they were more easily kept within bounds and 

 their fewer and larger leaves made it easier to locate the larvae. Ovi- 

 position records were ' obtained by confining freshly emerged adult 

 flies, either pairs or single females, in inverted lantern chimneys placed 

 over these plants and closed at the top by pieces of cheesecloth held 

 in place by rubber bands. A few records, also, were obtained in 

 larger cloth-covered rearing cages. The adults were transferred, 

 usually every day or two, to fresh plants, in order to ascertain more 

 exactly the incubation period of the eggs. These were examined 

 every morning, and the larvae, as they hatched, were transferred to 

 other plants, or, if not required, were destroyed, because, if left on 

 the same plant, they frequently ate into the unhatched eggs. Sev- 

 eral larvae were ordinarily placed on each plant, since the young ones 

 sometimes fell off and were lost. There was a great regularity in 

 the early molts, making them easy to keep track of, and as the larvae 

 grew larger they were moved to new plants, only one or two being 

 kept on a plant. The larvae showed little inclination to wander, and 

 the plants were left uncovered on the shelves. As the larvae neared 

 the final stage, however, a lantern chimney was inverted over the 

 plant, and several pithy stems, such as those of wild mustard, were 

 placed in the chimney for them to burrow into. However, if the 

 larvae were discovered just after the last molt, they were transferred 

 to cotton-stoppered glass vials with one of the stems, since they 

 sometimes burrowed into the ground instead of into the stems. Pu- 

 pation was observed by splitting open the stems a few days after the 

 larvae had entered. This did not a"ffect them at all, except that dur- 

 ing the hot, dry weather they showed a tendency to dry up and die, 

 either before or after pupating. This was avoided by moistening the 

 cotton plugs and by keeping the vials in an improvised moist chamber 

 made of a large lard can. Experiments showed that this did not 

 materially affect the pupal period; indeed, the presence of moisture 

 is undoubtedly a normal condition, since the dead stems in which the 

 larvae naturally pupate must lie on or near the moist soil in which 

 the food planl is growing. These careful laboratory studies were 

 supplemented by observations in the field, covering chiefly the 

 oatura] habits of i lie insect. 



NUMBER OK (^RNKRATIONS. 



Four general ions of the dork I'ldse-worm have been distinguished. 

 The generations overlap and become somewhat confused, the reason 

 being apparenl when ii is noted thai the emergence from overwinter- 

 ing larvae of the spring brood of adult flies in L915 covered a period 

 of even weeks, namely, from April L3 to June 1, with a single emer- 

 gence <>n June 16. In rearing this in ect, it was planned f<» carry a 

 genealogical series through the season. This attempt was only par- 

 i tally uccessful, since the hot , dry weather of the latter part of Julj 



