434 



ducts passes backwards and communicates with the two large 

 salivary glands (fig. 142), which reach back to about the 

 fourth body segment. The distal extremity is drawn out into 

 a narrow prolongation of the main salivary gland. 



Running along the sides of the midgut, and opening into 

 it posteriorly, is a pair of long moniliform tubes, the haepatic 

 c^aeca.^ A third such tube, shorter, however, than these, 

 opens into the midgut behind, and passes backwards over the 

 rectum (fig. 140). The two lateral caeca have been observed 

 by various investigators in the larvae of chalcid wasps. The 

 third median caecum has not, so far as I am aware, hitherto 

 been described. But the interpi;etation which is always placed 

 on them is quite erroneous. They are regarded as malpighian 

 tubes, but have in reality an entirely different function. For 

 this interpretation see, for example, M. Havilarid (1920, 

 1921). They are digestive glands, and in both structure (as 

 will be seen later) and embryonic development, differ entirely 

 from true malpighian tubes. The latter are ectodermal in 

 origin; the structures that occur in the larva of Nasonia are 

 outgrowths from the endodermal midgut. Furthermore, they 

 have no opening to the exterior during larval life, and empty 

 their secretions into the blindly-ending midgut. 



The necessity for such large haepatic caeca is clear when 

 we consider the rapidity of feeding during larval life; the 

 salivary glands are quite unable to cope with so great a 

 quantity of food, and well-developed haepatic caeca are but 

 to be expected. The secretion is poured into the posterior 

 part of the midgut, and the mixing with the engulfed food 

 is the result of a remarkable forward peristalsis which can 

 clearly be observed about once every fifteen to twenty seconds 

 in the feeding larva. A peristaltic wave travels along, from 

 behind, forwards, not only on the walls of the midgut, but 

 also on the body surface itself, and this brings about a per- 

 fect churning of the contents of the great food sac. 



Wlien now we look for the true malpighian tubes, we. 

 find that they are absent, and that the larva is entirely 

 devoid of excretory organs. And unless excretion occurs by 

 the cliffusion of ammonia through the integument, no excretory 

 activity goes on during active larval life. 



This fact, however, is less remarkable than it may at 

 first sight appear to be. The larva is exceedingly sluggish, 

 and is feeding upon food which has approximately the chemical 

 composition of its own tissues. Practically the only energy 

 expended is that which is needed in growth, and it is con- 

 ceivable that the excretory products resulting from the neces- 

 sary protein deaminisation accumulate in the blood during 

 the five days of larval life. In the late larva numerous 



