NO. 14 INSECT ABDOMEN SNODGRASS 5 



adoral rather than preoral, since their posterior ends embrace the 

 stomodaeum and give rise to the stomodaeal musculature (Kennel, 

 1888; Evans, 1902). The first truly postoral coelomic sacs of the 

 Onychophora are those of the jaws (/) ; the next pertain to the oral 

 papillae (OF). 



Among the Arthropoda a teloblastic origin and growth of the coe- 

 lomic mesoderm occurs in a sufficient number of cases (in certain 

 Crustacea and Chilopoda) to suggest that it is the primitive method 

 of mesoderm formation in this group as well as in the Annelida and 

 Onychophora. The mesoderm bands extend into the procephalic lobes 

 of the head as in adult Annelida and Onychophora, and may become 

 here excavated not only by antennal (antennular) and preantennal 

 coelomic sacs (fig. i C, lAnt, Prnt), but also, according to the obser- 

 vations of Wiesmann (1926) on Caraushis morosus, by a pair of sacs 

 in the labral region {Lm) lying immediately before the mouth. The 

 preantennal and antennal sacs in this case, it should be observed, do not 

 intervene between the labral sacs and the mouth, and are therefore not 

 preoral ; they lie at the sides of the mouth and are hence adoral. The 

 labral sacs (or perhaps ocular sacs), on the other hand, are literally 

 preoral, being adjacent to each other before the mouth. We may con- 

 clude, therefore, that in the Arthropoda, as in the Annelida, the coe- 

 lomic mesoderm potentially surrounds the blastopore by the union of 

 the mesoderm bands before the mouth, and that the antennal, pre- 

 antennal, and labral coelomic sacs, when present, lie on adoral and 

 preoral radii centering in the mouth (fig. i C). Whether these radial 

 cavities of the mesoderm are to be called " somites " or not becomes 

 largely a matter of definition. If the presence of a pair of cavities in 

 the mesoderm associated with a pair of external organs and their 

 ganglia is taken to define a segment, we may claim that the acron 

 of the arthropods is potentially a segmented region. On the other 

 hand, if a segment is conceived to be an independently movable sec- 

 tion of the trunk, then there is no segmentation of the arthropod trunk 

 anterior to the somite of the second antennae, shown either in the 

 embryonic procephalon or in the corresponding part of the adult head 

 capsule. According to Sollaud (1923), the first intersegmental groove 

 in the embryo of Palaemoninae separates an anterior head region, 

 or acron, from the first somite, which latter carries the second 

 antennae. The crustacean acron, or preoral region of the embryonic 

 head bearing the eyes and the first antennae, Sollaud believes, cor- 

 responds with the annelid prostomium. If this be true, it must be 

 admitted that the prostomial region of an onychophoran (fig. i B) 



