544 W. L. TOWER, 



and delicate cell walls are formed. This results in the development 

 of a protoplasmic body with numerous nuclei separated by partial 

 cell walls (PI. 18, Fig. 48). In this mass there develops a delicate 

 and much coiled tube with delicate structureless intima of chitin (?), 

 the whole appearing exactly like the development of the tracheoles of 

 the Lepidoptera as figured and described by Gonin (1894). The 

 tubule arises at the place where the mass of cells is in contact with 

 the intima of the tracheal trunk as a short, thick, cylindrical structure 

 with one end abutting directly against the intima and the other con- 

 stantly advancing into and eventually nearly obliterating the proto- 

 plasmic mass. When fully formed, the tubule, much coiled and con- 

 voluted, and the degenerating nuclei are formed in place of the cellular 

 body with which the structure started (PI. 18, Fig. 48 tr.cl). 



The development of the tubule from the protoplasm I have not 

 been able to observe. The process involved is certainly peculiar and it 

 is almost impossible to get any direct evidence upon it. It is certain 

 for two reasons that the tubules are not empty and that they are 

 not filled with air: 1) There is no way for air to get into the tubule 

 except by penetration through the intima of the tracheal trunk; 

 2) In freshly killed material which has not been treated with alcohol 

 or any reagent that would remove air the tracheoles do not show 

 any traces of air in them, but are filled with a highly refractive fluid. 

 After the tissue has been treated with alcohol and the processes 

 necessary for sectioning, these tubules are found to be empty so that 

 in life they evidently contain some substance which is soluble in and 

 removed by the reagents used in preparation, and is removed by 

 evaporation or drying when the tubule becomes functional. 



When the tubule is fully developed and the time arrives for it 

 to enter the wing, it unrolls and travels towards and into the wing 

 cavity. The cause of this unwinding is not easy to discover. To some 

 extent it may be aided by the entrance of the permanent tracheae 

 which begin to bud out from the dorsal tracheal trunk and enter the 

 wing at about this time. It is also almost certain that the tracheoles 

 are not unrolled until they become functional. It has been suggested 

 by Gonin (1894) and Mercer (1900) that the entrance of air into 

 them is the cause of the unwinding of the tracheoles. It is not easy, 

 however, to comprehend the manner in which the air is able to cause 

 this unwinding of the ball-like mass of tracheole. 



In the Lepidoptera according to Gonin (1894), the tracheoles do 

 not become functional until an ecdysis has been passed through, but 



