CLASSIFICATION OTT THE ANIMAL KINGDOM. 217 



In 1852 I gave full reasons for believing that the branchial sac of 

 the Ascidian " represents, not the gill of the Mollusk, but the per- 

 forated pharynx of ^m/?7woa7ws "* ; and I described the develop- 

 ment of the muscles of the tail in the larval Ascidian as " closely- 

 resembling that of the muscles of the Tadpole ;" but in the absence 

 of any sufficiently detailed knowledge of the development of the 

 embryo of either the Ascidian or of Ampliioxus, it was impossible 

 to know what weight ought to be attached to these resemblances ; 

 and it was not until the publication of the memoir of Kowalewsky 

 on the development of Ampliioxus that their real significance 

 became manifest. 



In this animal, in fact, yelk- division gives rise to a vesicular 

 Morula, which becomes provided with an alimentary cavity by in- 

 vagination, and with a cerebrospinal axis by the development of 

 laminae dorsales and the invagination of the corresponding portion 

 of the epiblast, as in other Vertebrata. 



The branchial clefts are secondary perforations of the body- 

 wall and pharynx ; and the protovertebrae and notochord are de- 

 veloped, as in Annelids and Arthropods, out of a mesoblastie layer 

 situated between the epiblast and hypoblast, and therefore in the 

 blastocoele. But one of the most important points made out by 

 Eowalewsky is, that the branchial clefts at first open externally — 

 and that they only acquire their anomalous position in the adult 

 by the growth over them of two laminae of the body-wall, which 



Society of Edinburgh, May 3rd, 1841, and published in vol. xv. of the "Trans- 

 actions' of that Society. "Viewed as an entire animal, the Lancelet is the 

 most aberrant in the vertebrate subkingdojn. It connects the Vertebrata, not 

 only to the Annulose animals, but also, through the medium of certain symme- 

 trical Ascidice (lately described by Mr. Forbes and myself), to the Molluscs, 

 We have only to suppose the Lancelet to have been developed from the dorsal 

 aspect, the seat of its respiration to be transferred from the intestinal tube to 

 a corresponding portion of its skin, and ganglia to be developed at the points 

 of junction of one or more of its anterior spinal nerves and inferior branch of 

 its second pair, to have a true annulose animal, with its peculiar circidation, 

 respiration, generative organs, and nervous system, with supra-cesophageal 

 ganglia, and dorsal ganglionic recurrent nerve." 



With every desire to give credit for sagacity where it is due, I think it is 

 obvious from this passage, and from the fact that Goodsir denied the existence 

 of the branchial clefts, or even of the abdominal pore, in AmpMoxus, that he 

 had no conception of its true morphological relations, and no valid grounds for 

 the hint which he throws out. 



* Report of the Belfast Meeting of the British Association, 1852. Trans- 

 ctions of the Sections, pp. 76, 77. 



