GIGANTOCYPRIS MULLERI 199 



are quite distinct at the upper end of the muscles disappear towards the lower ends so 

 that the myofibrils form a continuous bundle. Now these gradually spread out and run 

 through the substance of several ectoderm cells before reaching the cuticle. All these 

 morphological points suggest a persistent continuity between the muscle and the ecto- 

 derm cells and are in agreement with the idea of the muscle itself being ectodermal. 



However, my strongest point in support of this view, I consider, comes from a con- 

 sideration of the possible evolution of these muscles. We know, from the recent study 

 of crustacean embryology, that the fate of the coelomic sacs is constant throughout the 

 group. Dorsally they creep up on either side of the gut to form the heart and then, 

 after giving off the dorsal longitudinal muscles, they collapse to form the pericardial 

 floor. Ventrally they give rise to a mass of mesoderm from which develops the ventral 

 longitudinal series and sundry other muscles which may run in any direction and may 

 be associated with movements of limbs on the segments. The dorsal and ventral longi- 

 tudinal series are concerned only with extension and flexure of the body as a whole. 

 Moreover, the dorsal longitudinal series are the only muscles above the pericardial floor. 

 In the development of these muscles the cells forming the upper and outer wall of the 

 sac multiply and arrange themselves longitudinally. In this way one continuous longi- 

 tudinal muscle is formed. This may subsequently elaborate so that parts of it become 

 twisted in relation to other parts as, for instance, in the trunk musculature of a Decapod. 

 But always, whether the musculature is twisted or not, it runs from one segment to 

 the segment behind. Now in Gigantocypris we have just the reverse. The circular 

 muscles, even in their embryonic stages, run in a plane transverse to the body. There 

 is never any obliquity or, during development, any twisting round of originally longi- 

 tudinally placed muscle cells. If now the circular muscles are mesodermal they must 

 have evolved from the longitudinal series, for as I have emphasized this series is the 

 only series above the pericardial floor and is present throughout the Crustacea, or even 

 Arthropoda. This being so, there must have been stages leading up to these present 

 conditions where the mesoderm cells swing round from their original longitudinal 

 position to their present transverse position. Now, as I have just pointed out, there is 

 no trace during development of a rotation of mesoderm muscle cells, but more par- 

 ticularly it is difficult to see what advantage the initial stages in such an evolutionary 

 step would have been. 



In considering my suggestion that the circular muscles are derived directly from the 

 ectoderm it must be remembered that the fact that the dorsal body wall of the trunk 

 region is soft and flexible is just as characteristic of the Ostracods as is the possession 

 of a bivalve shell. The two characters are probably correlated and evolved together from 

 an ancestral form in which the tergal exoskeleton of the trunk region was rigid in the 

 typical crustacean manner. But it was to this dorsal skeleton that the segments of the 

 dorsal longitudinal muscle were attached. Now as the dorsal chitin gradually softened 

 this would cause the support for these attachments to disappear. A muscle cannot work 

 unless it has something rigid to pull against, and hence it can be deduced that as the 

 rigid dorsal chitin softened a compensation must have developed to take its place. 



