[419] _Zeit. f. wiss. Zool._, xx., 1870. For a recent view of the relation of the Enteropneusta to the Echinoderma, see J. F. Gemmill, _Phil. Trans._ B., ccv., pp. 213-94, 1914. [420] In a series of papers published in 1884-6, the speculative results being discussed in his memoir on "The Ancestry of the Chordata," _Q.J.M.S._ (n.s.), xxvi., pp. 535-71, 1886. [421] Reprinted in _Zoological Articles_, London, 1891. [422] "Die Enteropneusten des Golfes von Neapel," _Fauna und Flora des Golfes von Neapel_, Monog. xviii., Berlin, 1893. [423] See Macbride, "A Review of Prof. Spengel's Monograph on Balanoglossus," _Q.J.M.S._, xxxvi., 1894, and "The Early Development of Amphioxus," _Q.J.M.S._, xl., 1898. CHAPTER XVI THE GERM-LAYERS AND EVOLUTION In his papers of 1866 and 1867 Kowalevsky had remarked upon the widespread occurrence of a certain type or fundamental plan of early embryonic development, characterised by the formation, through invagination, of a two-layered sac, whose cavity became the alimentary canal. This developmental archetype was manifested in, for instance, _Sagitta_,[424] _Rana_,[425] _Lymnaea_,[426] _Astacus_,[427] _Phoronis_,[428] _Asterias_,[429] _Ascidia_,[428] the _Ctenophora_,[428] and _Amphioxus_.[428] He noticed also that the invagination-opening often became the definitive anus. Further instances of this mode of development were later observed by Metschnikoff[430] and by Kowalevsky[431] himself, but it was left to Haeckel to generalise these observations and build up from them his famous Gastraea theory. This was first enunciated in his monograph of the calcareous sponges,[432] and worked out in detail in a series of papers published in 1874-76.[433] Haeckel maintained that the "gastrula" stage occurred in the development of all Metazoa, and that it was typically formed, by invagination, from a hollow sphere of cells or "blastula." This typical formation might be masked by cenogenetic modifications caused chiefly by the presence of yolk. The gastrula stage was the palingenetic repetition of the ancestral form of all Metazoa, the Gastraea. From the Gastraea theory there followed at once two consequences, (1) that ectoderm and endoderm, invagination-cavity (_Urdarm_) and gastrula-mouth (_Urmund_ or _Protostoma_), were, with all their derivatives, homologous, because homogenous, throughout the Metazoa, and (2) that the descent of the Metazoa had been monophyletic, since all were derived from the ancestral Gastraea. Huxley's suggestion (_supra_, p. 208) that the outer and inner layers in Coelentera were homologous with the ectoderm and endoderm of the germ was thus fully confirmed and greatly extended.