448 



out of consideration the collateral evidence to be derived from the struc- 

 ture of Actmotrocha, but it will be evident to all who have read my 

 work on this form that its resemblances on the one hand, to Balano- 

 glossus, and on the other to Cephalodiscus, greatly strengthen the ho- 

 mologies put forward above. 



There can be no doubt whatever that the sub-neural gland of 

 Actitiotrocha is epiblastic (see p. 300). 'In the immediate neighbour- 

 hood of the mouth' does not accurately describe its origin (p. 300, 

 woodcut 1). It first appears well outside the mouth, in the epiblast 

 of the hood. If, as I understand Mr. H armer, this organ is to be com- 

 pared with the 'Eicheldarm' of Balanoglossus , the homology of the 

 latter with the noto chord of the Chordata is untenable. 



On the other hand, the sub-neural gland of Actinotroclia is 

 exactly comparable in form, function and relationships to other organs, 

 with the sub-neural gland of Cephalodiscus (Mr. Harmer's noto- 

 chord) and the structures I have compared with it in Balanoglossus. 

 PL XXI fig. 42 and PI. XXIII fig. 2 are worthy of comparison, as 

 regards the sub-neural sinus and proboscis pores. 



We now come to Mr. Harmer's third formula: 



3) 'The proof of the homology of the so-called 'paired notochords' 

 of Cephalodiscus and Actinotrocha with the notochord of higher C bor- 

 data is not convincing.' 



I cannot here reproduce the arguments which were given in favour 

 of this homology, but am not surprised that at present Mr. Harmer 

 is not convinced. The objections he puts forward shew that, possibly 

 because I did not make my views sufficiently clear , be has entirely 

 misunderstood my standpoint. 



This is seen most clearely in his reiterated assertion of the essen- 

 tial connexion of Balanoglossus in the establishment of the homology. 

 I do not regard Balanoglossus as forming 'a transition to the C bor- 

 data' and the existence of this group has not the slightest direct 

 bearing upon the homology of the 'paired notochords'. The claims of 

 these organs in Cephalodiscus and Phoronis to notochordal value stand 

 entirely upon their identity in structure and origin with the chordate 

 organ and these would be of as strong a nature even if such a group 

 as Enteropneusta were unknown to science. 



If Mr. Harmer will refer to p. 334 of my work, as well as to the 

 accompanying tree, he will see that my idea was a suggested derivation 

 of the Euchordata directly from an Actinotrochan-like ancestor of 

 the Archi-chorda and not through the H e mie h or da. The paired 

 pleurochords of this type have persisted in Cephalodiscus and in young 

 Phoronis but have fused together and extended pre-orally in the case 



