208 THE EMBRYOLOGY OF THE HONEY BEE 



the insects as any existing group of the arthropods, the mesoderm, 

 shortly after its formation, becomes broken up into segments, the 

 mesodermal somites, and these in turn give rise to a series of 

 paired sacs, the coelomic sacs, a pair of these corresponding typi- 

 cally to each segment. Each of these sacs sends out a hollow 

 prolongation into the appendage of the corresponding side of the 

 segment. The presence of paired coelomic sacs is probably primi- 

 tive since it is also characteristic of the annelid worms. A pre- 

 cisely similar condition is found in the trunk region (thorax and 

 abdomen) of Lepsima (Heymons 1897) and several members 

 of the Orthoptera (see Heymons 1895). ^ n au * these forms the 

 walls of the mesodermal sacs are thin and the various derivatives 

 of the mesoderm are not formed in situ, as in the honey bee and 

 many other forms, but from infoldings of the walls of the sacs. 

 In Forficula (Heymons 1895) the coelomic sacs have thicker 

 walls than in the forms previously mentioned and their deriva- 

 tives are formed directly from them, without the intervention of 

 folds. In the Coleoptera (e.g. Hydrophilus Heider 1889) a still 

 more modified condition exists. Here the coelomic sacs of the 

 trunk are thick-walled, and their lumen is relatively small, being 

 restricted to the extreme lateral region of the mesoderm, a 

 condition recalling that found in the honey bee, except that in 

 the latter the cavities are continuous longitudinally, constituting 

 the mesodermal tubes. In the muscids (Graber 1889) the cavi- 

 ties of the mesodermal somites are entirely absent, the mesoderm 

 being solid throughout. This is without question a highly special- 

 ized condition. 



In the majority of instances the cavities of the coelom are 

 formed secondarily by clefts which appear in the previously solid 

 mesoderm. Heider (1889) however has maintained that the 

 coelomic cavities are produced by a separation of the two layers 

 of mesoderm previously formed by the longitudinal folding of 

 the blastoderm (see p. 49). This view was strongly opposed by 

 Graber (1891) and has since not been confirmed by other investi- 

 gations on the development of the Coleoptera ; on the other hand 

 Carriere (1890) has found that the coelomic cavities in Chali- 

 codotna are formed in precisely this manner. 



In all cases there is a median layer of mesodermal cells connect- 

 ing the two lateral rows of mesodermal somites. 



