18 Journal New York Entomological Society. [Vol. xm. 



A". Conclusion. 



It has been shown in this paper that among the Coleoptera are to 

 be found examples of the simplest type of wing development known 

 to occur in the Holometabola, while this type differs from the de- 

 velopment of the wings in the Heterometabola mainly in that it is 

 held closely to the side of the body until pupation, and that in the 

 Scolytidae and in certain Bruchidas and Buprestidae, the development 

 of the wing takes place without any preliminary invagination of the 

 center of the disc or recession from the cuticle. It has been shown, 

 also, that there is a distinct system of tracheoles developed at the base 

 of the wing, the first appearance of which is coincident with the form- 

 ing of the wing disc, and that these tracheoles cannot be distinguished 

 from those that arise within the wing from the permanent tracheae, 

 during the prepupal period. 



I have shown conclusively that the wings do not arise from any 

 part of the spiracles of the mesothorax or metathorax, nor do the spir- 

 acles and the wings arise from the same place on the pleurum, even in 

 those insects in which the metathoracic spiracle degenerates. But 

 that the spiracles arise in the embryo in a different position from that 

 in which the wings arise, while in many insects the metathoracic spir- 

 acle does not degenerate. 



In those insects possessing tracheal gills, these gills are developed 

 in larval or pupal life and are temporary, adaptive structures that do 

 not persist in the adult, while there is no evidence whatever to show 

 that the wings have been derived from any such structure possessed by 

 the ancestors of the winged insects. 



On the other hand, the only conclusion that seems at all reason- 

 able, and the one to which the earliest stages in the formation of the 

 wing in all insects seems to point, is, that the wings have been derived 

 as lateral outgrowths or folds of the hypodermis of the pleurum or ter- 

 gum, or both. 



Methods. 



In preparing the material for these investigations several different 

 methods were tried. Among the fixing solutions, the best results were 

 obtained with a saturated solution of corrosive sublimate in ten per 

 cent, formalin, washed out with four per cent, formalin. Very good 

 results were also obtained by the use of Tower's No. 3, the formula 

 for which is as follows : 



