488 A. J. MASTEKMAN. 



of the work referred to^ yet tlie agreements are so little 

 emphasised, and the disagi-eements so dwelt upon, that any 

 reader who had not studied my work could almost suppose 

 that the total result was rather a refutation than a corroboi*a- 

 tion of the latter. However, Iked a's figures are so accurately 

 and carefully drawn that it is possible to explain some of the 

 more important differences. The first kind of discrepancy 

 is simply due to a difference of interpretation. For example, 

 he shows clearly in several of his figures (such as the series 

 fig. 59, a, h, c, d) the snbneural blood-sinus in even greater 

 detail than myself, but he prefers to label it as an artefact. 

 As the blood system of Actinotrocha is h^emocoelic, and 

 hence consists of mere sinuses or spaces between the three 

 primary layers, it is naturally possible to regard it in its 

 entirety as an artefact. The same remark applies equally to 

 the whole vascular system of Cepha Iodise us, the greater 

 part of that of Balanoglossus, and that of numerous larvfe. 

 ludeed, there is little to urge against the assertion that the 

 heemoccelic body-cavity of arthropods is an artefact. On 

 similar grounds he fails to corroborate the presence of the 

 perianal blood-sinus (one of the most conspicuous characters 

 of the larva), and of other parts of the blood system. Similar 

 considerations apply to his refusal to recognise the presence 

 of the epidermic pit (" neuropore ") and of the subneural 

 gland. Both of these are clearly figured in his sections with 

 exactly the relations indicated by myself, but he prefers to 

 regard them as artefacts. As in all my fully developed 

 larvae, alive or dead, I have never failed to find all these 

 organs constant in position and in structure, I can hardly 

 agree that they are mere freaks of reagents. 



Ikeda suggests that the epidermic pit in front of the 

 ganglion is produced by contraction of muscles drawing 

 the ganglion backwards (!), and similarly he accounts for 

 the evident presence of the snbneural gland by an artificial 

 bulging forwards and outwards of the part of the hood im- 

 mediately anterior to it (!). Surely the inechanical con- 

 ditions induced in each case would tend to exactly the 



