ANATOMY AND PHYSIOLOGY 45 



number of segments entering into the composition of the insect 

 head has been a difficult problem. As no segment bears more 

 than one pair of primary appendages, there are at least as 

 many segments in the head as there are pairs of primary 

 appendages. On this basis, then, the antennae, mandibles, 

 maxillae and labium may be taken to indicate so many seg- 

 ments; but in order to decide whether the eyes, labrum and 

 hypopharynx represent segments, other than purely anatom- 

 ical evidence is necessary. The key to the subject is furnished 

 by embryology. At an early stage of development the future 

 segments are marked off by transverse grooves on the yentral 

 surface of the embryo, and the pairs of segmental appendages 

 are all alike (Fig. 194), or equivalent, though later they dif- 

 ferentiate into antennas, mouth parts, legs, etc. Moreover, the 

 nervous system exhibits a segmentation which corresponds to 

 that of the entire insect ; in other words, each pair of primitive 

 ganglia, constituting a neuromere, indicates a segment. Now 

 in front of the oesophagus three primitive segments appear, each 

 with its neuromere (Fig. 55) : first in position, an ocular seg- 

 ment, destined to bear the compound eyes ; second, an antennal 

 segment; third, an intercalary (premandibular) segment, 

 which in the generalized orders Thysahura and Collembola 

 bears a transient pair of appendages that are probably homol- 

 ogous with the second antennae of Crustacea. In the adult, 

 the ganglia of these three segments have united to form the 

 brain, and the original simplicity and distinctness have been 

 lost. The labrum, by the way, does not represent? a pair of 

 appendages, but arises as a single median lobe. Behind the 

 oesophagus, three embryonic segments are clearly distinguish- 

 able, each with its pair of appendages, namely, niandibularjna.v- 

 illary and labial. Finally, the hypopharynx, or rather a part of 

 it, claims a place in the series of segmental appendages, as the 

 author has maintained; for in Collembola its two dorsal con- 

 stituents, or superlinguce, develop essentially as do the other 

 paired appendages and, moreover, a superlingual neuromere 

 (Fig. 55) exists. The four primitive ganglia immediately 



