26 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I46 



vealed the presence of small, paired, temporary cavities in the meso- 

 derm of the antenna! region of the head, and another pair in the 

 preantennal region. Some embryologists insist that any pair of 

 coelomic sacs must represent a segment. They contend, therefore, 

 that primitively the embryonic head lobe of insects contained botTi 

 an antennal segment and a preantennal segment, thereby making 

 six primary segments in the definitive head in addition to a small 

 anterior prostomial region bearing the labrum. [The maximum num- 

 ber of segments in the insect head, based on these coelomic sacs, is 

 nine according to Janet; four of these would be in front of the 

 antennal segment, if the antenna does indeed respresent a segment.] 

 The contention is logical if we accept the premises. Coelomic sacs 

 are spaces in the mesoderm for the accumulation of waste products 

 of metabolism, and most of them have ducts leading to the exterior. 

 The outer walls of the sacs form the longitudinal muscles that deter- 

 mine the segmentation of the ectoderm. Where there is no ectoder- 

 mal segmentation, as in the embryonic head of modern arthropod 

 embryos may we not question that coelomic sacs are always ac- 

 companied by ectodermal segments? Those of the embryonic head 

 may have purely a physiological function (and a transitory one since 

 they are not carried over to later stages of the insect). The reported 

 presence of coelomic sacs in the labrum is particularly difficult to 

 account for since few morphologists regard the labrum as repre- 

 senting a segment. The actuality of an antennal segment and a 

 preantennal segment in the primitive head of insects may, therefore, 

 be doubted, but not outright denied. 



A theory of head segmentation promulgated a few years ago 

 caused much confusion by its sensational claim that the tritocerebral 

 lobes of the brain are the ganglia of a "labral segment," because the 

 labral nerves are connected with them (Ferris). This idea was based 

 on observation that the endings of body nerves remain in the segment 

 of their origin even after their central ganglion has been transposed 

 to another segment, and that thus one may identify the segment of 

 the ganglion. This generalization is true for motor nerves which 

 arise from the ganglia and grow outwardly to the part which they 

 will innervate. The labral nerves in question, however, are sensory 

 nerves arising in the labral epidermis and growing in to the trito- 

 cerebral ganglia. The origins of sensory nerves do not identify the 

 segment of the ganglion to which the nerves go, and nothing indi- 

 cates that the labrum is a head segment. We have here an excellent 

 example of how revolutionary ideas can be drawn from logical 

 reasoning based on false premises. 



