282 COMPARATIVE ANATOMY CHAP. 



pole of the larva a small depression of the ectoderm becomes connected with the 

 enteric tube. The permanent mouth and the stomodteum thus arise. The primitive 

 mouth closes. The enteric tube becomes a solid strand, which continues to grow till 

 it reaches the closed primitive mouth, i.e. the posterior end of the body. An enteric 

 cavity does not again appear till a later stage. On the ventral side, a little in 

 front of the posterior end of the body, an anus forms. The postanal intestine 

 degenerates, its 2 visceral layers of the mesoderm forming the septum of the caudal 

 segment. From the 4 genital cells the testes and ovaries develop. We thus see 

 that in Sagitta the mesodermal organs have a double origin : first, 2 (afterwards 4) 

 large endoderm cells lying at the base of the archenteron, which form the rudiments 

 of the sexual glands ; secondly, the whole middle and oral epithelial wall of the 

 archenteron of the ccelogastrula ; this represents a germ zone, continued at the 

 edge of the blastopore into the ectoderm, and at the aboral portion of the larva 

 into that part of the wall of the archenteron from which the definitive intestine is 

 produced. This germ zone forms at an early stage two lateral cceloine sacs as a 

 consequence of the formation of the folds above mentioned ; the cavities of these 

 sacs produce the body cavity, and their walls the endothelium of the body cavity, 

 and probably also the musculature of the body wall. 



Two hollow lateral coelome sacs of the archenteron appear at an early stage in 

 one of the Brachiopoda, Argiope, much in the same way as do those in Sagitta ; these, 

 constricting themselves off from the intestine, are said to form the body cavity and 

 the mesoderm. The development of the mesodermal organs has, however, up to the 

 present time been insufficiently observed. 



Various theories have been put forward as to the phylogenetic significance of the 

 different processes of development to which the mesodermal organs owe their rise 

 ontogenetically, all these theories resting upon the assumption that the ontogenetic 

 process exactly repeats, sometimes in one point sometimes in another, the phylo- 

 genetic development. These theories rest upon weak foundations as long as com- 

 parative anatomy knows of no series of animal forms which shows us the gradual rise 

 of the mesodermal organs in a manner similar to that seen in the successive onto- 

 genetic stages of development. It is very doubtful whether the whole mesoderm, 

 except the nervous system, can be derived phylogenetically from such simple organs 

 as the ccelome sacs, or from cell groups such as the polar cells of the mesoblast, 

 once present in simple gastrula-like racial forms. Observations are increasing in 

 number which tend to show that ontogenetically also there is no single rudiment of 

 the whole mesoderm, but rather several rudiments for the different mesodermal 

 organs. Our review of the history of development of the worms supports this latter 

 view. It may perhaps in time be established that the manner in which the various 

 mesodermal organs appear in the Cnidaria moving from their places of formation, the 

 ectodermal body epithelium and the endodermal enteric epithelium, into the deeper 

 parts of the body wall, is essentially the same as that in which the mesodermal organs 

 originally arose in the ancestors of the Platodes and the Vermes. The ontogenetic 

 development of the mesoderm would then represent this process very much abbreviated 

 and localised, pushed back to very early stages. If, in the Hirudinea (in Clepsine at 

 least) and in Lumbricus as opposed to the other Annulata, the ventral chord does not 

 arise in situ in the ectoderm but is formed by two blastomeres, the neuroblasts, which 

 arise at an early stage, it is difficult to see why the polar cells of the other mesodermal 

 organs (nephroblasts, mesoblasts, etc.) should not represent similar early developing 

 condensed and localised rudiments. And why should not these different rudiments 

 themselves be pushed back to, and localised and condensed in, a rudiment such 

 as the early developed primitive mesoderm cells or zones 1 In the Polyclada we 

 see at an 8-micromere stage (Fig. 94, p. 125) the rudiment of the whole ectoderm, 







