206 PARASITES OF GIPSY AND BROWN-TAIL MOTHS. 



persistently remain in the uppermost recesses of the cage and refuse to 

 come down. 



All in all, the disadvantages were so many, in proportion to the 

 advantages, and these latter were so largely imaginary in point of fact, 

 as to result in the decision to discontinue the use of the large cages 

 entirely in 1908. 



The cage figured by Mr. Townsend (PI. XV) was, however, an 

 innovation in several respects. It was built independently of any 

 tree which should serve as food for the inclosed caterpillars, but 

 these caterpillars were confined within certain restricted limits and 

 exposed to the attack of the tachinid flies at one and the same time 

 by the use of the open " tanglef ooted " tray. Here a most distinct 

 advantage was gained. The floor of trodden earth (subsequently 

 replaced by cement) effectually prevented the entrance of numerous 

 insects which were formerly uninvited guests and thereby removed 

 another serious disadvantage. An arrangement of double doors 

 and wire-screened vestibule prevented the untimely liberation of 

 the flies, and there were no longer so many inexplicable disappear- 

 ances. The fact that the top of the cage was flat instead of being 

 extended into the gable tended to keep the flies somewhere more 

 nearly where they were wanted. In short, there were a great many 

 advantages possessed by the new cage which were not possessed by 

 the old, and there was some justification for considering it good. 



In the meantime Mr. Burgess, who had taken over the Calosoma 

 work in the fall of 1907, had developed the out-of-door cage along 

 totally different lines, making it into nothing more than an out-of- 

 door insect ary (PL XVI), in which were conducted practically all of 

 his numerous and varied investigations. It had seemed in 1907 as 

 though the only one among the numerous imported insects which 

 had done at all well in the out-of-door cages as then used had been 

 the Calosoma, but the success attending their use for the rearing of 

 this insect was so soon and so overwhelmingly eclipsed by the success 

 which attended the use of small individual cages for single pairs of 

 the beetles or individual larvae as to render the advisability of their 

 discontinuation for this purpose emphatic. 



Some attempt was made to use the tachinid cage in 1909, but not 

 to the extent to which it had been used the previous year. Late 

 in the summer of 1909 reproduction experiments with small numbers 

 of various species of tachinids were undertaken by Mr. W. R. Thomp- 

 son, who used cages constructed after the familiar Riley type, but 

 covered entirely with coarse fly screen. (PL XVII, fig. 1.) He 

 succeeded in much of that which he undertook to do, and in 1910 

 continued the use of this type of cage, for a part of a quite extensive 

 series of most interesting and successful experiments, but he also 

 used a much smaller cage consisting of a wire-screen cylinder (PI, 



