242 haeckel's anthropogenie. 



Norway, and wliicli, at one period of its life at any rate, would seem 

 to be most nearly related to the flagellate infusoria. 



The gastrsea, the ancestral form of all animals except protozoa, 

 possessed a simple, primitive intestine, with a primitive mouth, and 

 a body- wall formed of two layers, ectoderm and entoderm. The 

 gastrula stage of all metazoa repeats this in ontogenesis, and the 

 nearest living representatives are to be sought among the simplest 

 calcareous sponges, the asconidse. 



From the gastragadse development diverged in two lines : one of 

 these, radiate, giving rise to sponges and coslenterata, the other, bilate- 

 ral, to the vermes. It is among the latter that we must look for the 

 ancestors of man, from which group the vertebrate bilaterality has 

 been inherited. The recent vermes are divided into acoelomi and 

 coelomati, according to the absence or presence of a body-cavity. The 

 phylogenetic relationships of the acoelomi are sufficiently patent; but 

 intermediate forms must have existed between the gastrgeadse and such 

 highly organized forms as the recent turbellaria. These intermediate 

 forms (necessarily without a body-cavity) may be called archelminthes, 

 and from that group the accelomi and the coelomati diverge. 



The next stage must be looked for among .the archelminthes, 

 may be called prothelmis, and must have resembled a low turbel- 

 larian in the shape of intestine, absence of anus, possession of excre- 

 tory canals, hermaphroditism, &c. Prothelmis represents only one 

 of a very long row of forms which must have connected the gastrseadse 

 with the coelomati. The passage from prothelmis, with four secondary 

 germ-layers, as many turbellaria have, to a coelomatous form, is easy 

 to understand, and the formation of a body-cavity would be correlated 

 with a number of other important organological changes. The living 

 coelomati only represent the terminal twigs of an enormous tree, and 

 the only forms among them which are closely related to the vertebrata 

 are the tunicata. This resemblance, however, is to be observed 

 chiefly in development (and in such a persistent larval form as appen- 

 dicularia), and we must thus consider the tunicata and vertebrata to 

 be related merely by descent from a common ancestral group, the 

 chordonia, characterized by the possession of a chorda. 



Just as in the acoelomi, so in the coelomati, there must have been 

 a long row of intermediate forms which led from prothelmis to 

 chordonium, and one such stage may be easily arrived at and may be 

 called the scolecida. This seventh stage is characterized by the 



