4 
Journal of Agricultural Research v 0 i. xxviii, no. i 
third instar it is much larger and presents a clear, circular, windowlike area 
(PI. 1, J, n) in the outer side of its base, directed outward and forward. This 
area is connected with the body wall (PL 1, I, BW) just ectad of the base of the 
external lobe by a short' cuticular tube ( v) which ends in a circular scar (m) on 
the body wall. The same structure is repeated in the prothoracic spiracle of the 
pupa (PI. 2, A) and we shall see that it is characteristic also of the posterior 
spiracles of the second and third larval instars. It is to be explained as follows: 
Each spiracle is a newly formed structure developed, not within that of the 
preceding instar, as is usual in other cuticular structures, but from the inner 
side of the base of the stigmatic chamber. At each molt the old spiracle is 
cast off and the lining of the stigmatic chamber, together with the tracheal 
intima, is drawn out attached to it through the channel formed by the new 
cuticle of the old chamber. The latter then shrinks, as the new chamber becomes 
functional, and is reduced to the cuticular tube (v) connecting the base of the 
chamber with the new chitinous layer of the body wall, the closed ends forming 
the external and internal scars (m and n ). 
The successive formation of new spiracles in each instar of the fly larva has 
been described by De Meijere (SO) for such a large series of genera from Myceto- 
phila to Hydromyza that there can be no doubt that it is characteristic of all 
the amphipneustic Diptera. Though the form of the external spiracular lobe 
may vary from a simple projection to a many-fingered flap or to a long, slender 
tube, the fundamental structure is always the same. It has not been proved 
for all stages that the new stigmatic chamber is an outgrowth from the old, but 
Weismann (54) has shown in Corethra, and Wahl (49) in Eristalis, that the 
stigmatic chambers of the pupal spiracles grow out from the ends of the tracheae 
as solid masses of cells, proliferated from the tracheal epithelium, in which a 
lumen appears later. It would seem more reasonable that the chambers should 
be formed as secondary ingrowth from the body wall since they have no taenidia 
nor any other appearance of tracheal structure, and Haake (IS) says that they 
are thus formed in the larva of Trigona. De Meijere (SO) says that the inner 
walls of the chamber (“filzkammer”) are usually ciliated, and that in some 
forms they are braced by chitinous bars across the cavity. In Rhagoletis the 
chambers are unobstructed and their walls appear to be rugose but not ciliated. 
These dorsal spiracles of both the larva and the pupa are, in any case, secondary 
structures. They are developed in connection with the dorsal longitudinal 
tracheal trunks (PI. 2, G), while the primitive spiracles, temporarily closed in the 
maggot, belong to the lower lateral trunk. That special breathing structures of 
a similar nature occur in other orders is indicated by the description of a respira¬ 
tory organ in a beetle larva, Donacia, given by Boving (9). Here a hollow, 
apparently perforate, spinelike process is situated on the back of the eighth seg¬ 
ment over an end-chamber of a trachea, and contains a prolongation of the latter. 
At each molt the spine is cast off, bringing the tracheal intima with it, while a 
new spine is generated by the new cuticle and a new branch of the tracheal end- 
chamber grows out into it. 
The last body segment of the maggot is the largest (PI. 1, A, VIII), The 
posterior part of its upper surface is flat, slopes downward, and bears the posterior 
spiracles (PSp ). The anus (An) is situated on the rear part of its ventral surface 
between a pair of lobes. The stigmatic surface is surrounded by eight tubercles, 
two above and two below the spiracles, and two very small ones on each side. A 
fifth pair occurs between the substigmatic pair and the anal lobes. In the pupa 
the eighth pair of abdominal spiracles is situated on this segment, and the part 
bearing the anus is differentiated as a small segment which becomes the tenth of 
the adult. The terminal segment of the larva, therefore, is evidently a combina¬ 
tion of the eighth, ninth, and tenth abdominal segments. 
