123 



general results at which I have arrived in regard to the morphology 

 of the endostyle and other organs in the Chordata. The results upon 

 all the points which I have studied have proved so harmonious with 

 one another that I have been able to synthesise them into a consistent 

 theory of chordate phylogeny of which I here briefly sketch the out- 

 lines. The complete evidence for this view will shortly be published 

 in the Quarterly Journal of Microscopical Science. 



Stated shortly , my theory postulates a common ancestor for the 

 Echinoderma, the Enteropneusta and the Chordata. This ancestor was 

 bilaterally symmetrical and had the external appearance of a young 

 Auricularia larva, in which the blastopore (anus) is posterior and 

 terminal, and the sides of the body are margined by a complete circum- 

 oral ciliated ridge. At the apical or prae-oral pole was situated a 

 pair of pigmented ectodermal pits , resembling the eye-spots of Tor- 

 naria. These pits, together with the blastopore, lay within the dorsal 

 or aboral area circumscribed by the circumoral band. An adorai ciliated 

 band was also present and had the relations of the adorai band of 

 modern Auricularia larvae ; that is to say , it bounded the dorsal and 

 lateral margins of the mouth, and was prolonged ventrally into a loop 

 which penetrated some distance along the ventral wall of the oeso- 

 phagus. The Auricularia-\\^e ancestor possessed at least two pairs of 

 bilaterally symmetrical enterocoeles , of which the first pair communi- 

 cated with the exterior by water-pores. The central nervous System 

 of the pelagic ancestor consisted of an elongated nerve-ring which lay 

 exactly underneath the circumoral ciliated ring , and , therefore , like 

 the latter, enclosed both the apical plate and the blastopore. 



The Echinoderma were derived from this hypothetical ancestor by 

 a series of changes mainly correlated with the secondary assumption 

 of a radial symmetry. The right anterior enterocoele atrophied, in 

 the way recently described by Brooks and Field in certain Bipinnaria 

 larvae , the left anterior enterocoele persisted and retained its pore as 

 the definitive »dorsal pore«. The nerve-ring segregated itself from the 

 circumoral ciliated band, as described by Se mon in the development 

 of Holothurians, and moved ventrally so as to constitute the circumoe- 

 sophageal nerve-ring of the adult. The sense organs of the prae-oral 

 pole atrophied. 



The Chordata retained the primitive bilateral symmetry. The 

 circumoral ciliated ridges of the two sides approximated dorsally, and 

 fused along their entire length in the mid-dorsal line , enclosing both 

 the prae-oral sense-organs and the blastopore, in this way constituting 

 a canal, ciliated internally, and communicating with the gut by means 

 of the blastopore (neurenteric canal) . The ciliated ridges , upon this 



