544 Cc. H. RICHARDSON 
muscles present a very definite and regular series of longitudinal 
striae, but cross-striations are not visible at this stage of develop- 
ment. The hypodermal cells, on the other hand, do not possess 
such regular striae, but appear to be distinctly granular. To 
either side of the point of contact of the muscles, where muscular 
tissue is wanting, the hypodermal cells are of the same morpho- 
logical structure though less attenuated and with the Giemsa- 
eosin stain give the same reaction. Almost the identical con- 
dition has been found in the wing muscles of the Chalcidoid, 
Monodontomerus, by Berlese (09) and in the wasp Vespa by 
Torne (11). 
Large Tabanid larvae collected at Forest Hills, Massachusetts, 
proved to be excellent material for the study of muscle attach- 
ments. _ Sagittal sections 4u in thickness, stained with iron hema- 
toxylin-eosin show in a most conclusive manner the hypodermal 
attachment. At the point of insertion, the hypodermal cells are 
greatly elongated as in Spalangia. Their nuclei, which are not 
seen in figure 17, agree in structure with the adjacent hypodermal 
nuclei except that they are somewhat larger. A distinct line 
separates the muscles from the hypodermis, but I am not sure 
that the basement membrane is continuous between the muscle 
and hypodermis, although it can be traced in a little way on 
either side. Just within the hypodermis, where.the two muscles 
come together, there is a mass of quite homogeneous substance 
which I believe to be an intercellular deposit. 
Now, considering the nature of the dermal fibrillae which are 
seen in the dermis just above the elongated hypodermal cells, 
I believe that two facts preclude the possibility that they are 
fibrillar outgrowths from the muscles or hypodermal cells; first 
they do not take either the hematoxylin or eosin stains but 
remain slightly yellow-hyaline like the adjacent dermal substance; 
second, when sections are macerated in a strong aqueous solution 
of NaOH, they persist, even after the muscles have disappeared. 
With prolonged treatment, of course, the dermis disappears also. 
Figure 16 offers strong support to the view that the hypo- 
dermal fibrillae are not continuous with the muscle fibrillae. 
When sections of the tabanid larvae were treated gently with 
