54 
Psyche 
[June 
Longitudinal muscle fibers are absent. Four bulb canals , 
continuous with those of the tract and calyx, extend the 
length of the bulb and end blindly at the posterior extrem- 
ity. In the anterior half of the bulb they do not commun- 
icate with the bulb cavity, but are closed off by the ap- 
proximating chitin walls internally adjacent to them. 
Posteriorly, the walls diverge so that the canals communi- 
cate widely with the lumen when the transverse muscles 
are relaxed (pi. 7, fig. A) , but are closed off from the lumen 
when the muscles contract. As in the calyx and tract, the 
outer surfaces of the walls form the points of attachment 
of the transverse muscles. 
The bulb merges into the cylinder by a posterior con- 
striction and a termination of the bulb canals. The circular 
muscles are again reduced to a thickness of one to two 
fibers, and there is still no evidence of longitudinal muscle 
fibers (pi. 7, fig. B). Internally the cylinder consists of a 
simple chitinous tube surrounded by a single-cell layer of 
epithelium. At its posterior end, as it enters the midgut, 
the cylinder is seen in our preparations to fold upon itself 
in such a way as to become a compound tube (pi. 7, fig. C) . 
This is made possible by an extreme thinning of the intima 
in that area. At the end of the cylinder the intima is folded 
inside out and back upon the cylinder for a short distance; 
it is then turned inward again and envelops the cylinder 
as a third fold before it terminates at the midgut junction. 
The median intima shown in our figure is therefore actually 
the true intima folded upon itself. The significance of this 
rather complicated situation is that the tip of the cylinder 
is highly flexible, allowing the cylinder to be intruded far 
into the midgut as the cylinder valve (“knob” of previous 
literature) or to be pulled out of the cavity altogether as 
a straight, simple tube. Thus the relative lengths of the 
cylinder and cylinder valve, or those parts outside and 
inside the midgut cavity respectively, can vary consider- 
ably, and the cylinder valve can disappear entirely under 
some conditions. This phenomenon was first observed 
by Forel in Conomyrma pyramica (1929) ; we have seen it 
in several formicine genera, including Camponotus, Lasius, 
and Prenolepis. 
The midgut is of sufficient histological interest to deserve 
