NO. I THE INSECT HEAD — SNODGRASS 49 



theoretical in the absence of concrete evidence, and it is difficult to 

 visualize the embryonic head as having once consisted of individually 

 movable segments. It is easier to believe that temporary cavities can 

 occur in the preoral mesoderm without giving rise to segments. 

 Primitive coelomic cavities must have had some primary reason for 

 their formation, probably a physiological one. They usually set the 

 pattern for segmentation, but that they necessarily form segments 

 is just a convenient belief for supporting a theory when no segmenta- 

 tion is visible. DuPorte (1957) has well discussed the weakness of 

 evidence for segmentation in the preoral head region derived from 

 the presence of cavities in the mesoderm. 



It is true that Nelson (1915) describes protocerebral and deuto- 

 cerebral segments in the embryo of the honey bee, but it appears that 

 he refers to surface swellings over the ganglia ; no mention is made 

 of coelomic cavities in this region. Shaliq (1954) likewise finds no 

 coelomic sacs in the embryonic head of the sawfly Pteronidea rihesii, 

 and from the lack of any other evidence of segmentation he con- 

 cludes that the embryonic head is better interpreted as an unseg- 

 mented acronal lobe bearing the eyes, the antennae, and the labrum. 



The occurrence of cavities in the labral mesoderm should be some- 

 what embarrassing to the segmental theory regarding the rest of the 

 head lobe. Most embryologists do not accept the labrum as a seg- 

 ment, but they insist that the cavities in the following region denote 

 former segments. Yet the presence of paired mesodermal cavities in 

 the labrum appears to be as well attested as that of cavities in the 

 preantennal and antennal mesoderm. Paired cavities in the embry- 

 onic labral mesoderm have been described by Wiesmann (1926) in 

 Carausius, by Mellanby (1936) in Rhodnius, by Roonwal (1937) 

 in Locusta, by Eastham (1930) in Pieris, and by Miller (1940) in 

 Pteronarcys. The cavities, however, soon become disorganized and 

 their walls reduced to irregular cell masses. That the labrum con- 

 tains mesoderm in all cases is unquestioned, but the validity of the 

 labral cavities as true coelomic sacs is disputed by some writers, par- 

 ticularly by Manton (1928), who cites Wiesmann as the only one 

 who records the presence of labral sacs distinct from a pair of pre- 

 antennal sacs. More recently, however. Miller (1940) has described 

 in the stonefly Pteronarcys definite traces of cavities in the prean- 

 tennal mesoderm, as well as cavities in the labral mesoderm. 



To further support the claim of primary segmentation in the blasto- 

 cephalon, it will be argued that the presence of ganglia is in itself 

 evidence of segmentation. It is true, of course, that each pair of 



