8 Journal New York: Entomological Society. [Voj. xiii. 



from their nuclei. The manner of development of the tracheoles is 

 the same in all the beetles examined. Figs. 3 and 23 show an early 

 stage in their development. From these figures it will be seen that 

 they are formed as long curved and coiled tubes. As the cytoplasm 

 of the tracheal cells becomes used up in their development, these tubes 

 become looped around the nuclei and by the time that the wing is 

 well evaginated the greater part of the cytoplasm of the cells has been 

 used up and we find a mass of large tracheoles looped and coiled 

 around the nuclei (Fig. 7), nearly every nucleus being closely en- 

 folded by a tracheole. 



The tracheoles are present throughout the larval development of 

 the wing and probably degenerate some, time during the pupal period. 

 Fig. 17 shows a bundle of tracheoles extending into the base of a 

 wing in the prepupal period. Figs. 13, 15 and 16 show tracheoles 

 that have developed from the wing tracheae after their entrance into 

 the wing. I have not observed the earliest stages of the formation of 

 these secondary tracheoles, but it will be seen from these figures, 

 which were drawn from the same wing as Fig. 17, that there is no 

 difference in the structure of the primary and the secondary tracheoles. 

 In fact the two cannot be distinguished when in the wing. 



The simplest condition of- the tracheoles was found by Tower to 

 be in Coccinella bipunctata, in which a few tubules developed from the 

 tracheal trunk at irregular intervals, while the most specialized condi- 

 tion was found in the Buprestidre. My observations confirm those of 

 Tower that the most specialized condition of the tracheoles is to be 

 found in the Buprestidae, but I find, however, that as in T. plasto- 

 graphus and D. valens, the tracheoles begin to develop as soon as the 

 wing disc, and are fully formed, as a mass of tubes which are very 

 small next the wing disc, becoming larger further out, by the time 

 that the wing begins to evaginate (Fig. 19). 



The tracheoles in the Coleoptera are of a simpler type than is 

 found in the Lepidoptera as described by Gonin (1894), Mercer 

 ( 1900) and others. Mercer found the tracheoles beginning to develop 

 during the fourth larval stage and becoming functional after the next 

 moult, at the beginning of the last stage. In the Coleoptera, accord- 

 ing to Tower, they do not begin to develop until the wing is well 

 formed, during the last larval stage, becoming functional early in the 

 prepupal period and without a moult. I have not been able abso- 

 lutely to determine the time of their becoming functional in the bee- 



