230 PERCY SLADEN TEUST EXPEDITION. 



the oblique dark baud alluded to on p. 217, where it passes up from its anterior termination 

 in the ventral edge of tbe rostrum towards the edge of the nietapleure. The epidermal 

 band of the left fold is not seen until the level of the 4th gill-slit, just behind the 

 mouth ; its position corresponds completely with that described for A. valdivice. At the 

 h^vel of the 12th slit the two folds are seen to be approximately equal in size, although 

 ihey are, nowhere, perfectly symmetrical*. The epithelial band of the left fold is very 

 poorly developed. 



The pterygocoeles, as Goldschmidt points out, are here relatively minute ; but towards 

 the posterior end of the gill-region they become much larger than anteriorly {vide PL 15. 

 tig. 11). In front of the 6th slit or thereabouts I cannot discover them to exist at all ; 

 their absence anteriorly, if a fact, must have some bearing upon our views as to their origin 

 [vide infra, p. 253). Again I find no certain connection with the splanchnoccele anteriorly. 

 I looked for a connection posteriorly, with an equally unsatisfactory result. The 

 segmental connections which Van Wijhe (1906) notes in the Amphioxus larva do not 

 occur in Amphioxides ; these also were not regarded by their discoverer as certainties. 

 In both A. valdivice and A. pelagians a corner of the splauchnoccele, cut olf from the 

 rest by the transverse branchial muscle, can often be seen to project into the substance 

 of the metapleure ; it is quite distinct, ho^vever, from the pterygocoele. 



The Excretory Canals. 



I owe it to M. Lcgros that I am able to state the existence of excretory canals in 

 A. valdivice ; no doubt they exist in A. pelagicus also, although I have not yet been able 

 to see them there. Goldschmidt's failure to find them is not surprising, since their minute 

 size renders them more than inconspicuous, in the absence of exact knowledge as to the 

 right position for the search. I saw nothing of them until M. Legros kindly pointed 

 out to me the very similar tubules of the BrancMostoma larva ; with this aid their 

 identification became easy. 



The canals form a complete series, on the left side only, one in relation to each 

 gill-slit. Each is a very narrow tube lying between the splanchnopleure and the 

 thickened brancliial epithelium which lines the left-hand border of the gill-slit. Usually 

 the lumen of the tube can be seen in three successive sections of 6'6;u thickness. In the 

 most anterior (vide PI. 15. fig. 12) it is seen to abut upon the point where the thick branchial 

 passes into the thin external epithelium at the mouth of the slit. I have little doubt 

 that the tubule has an external opening here, though this could not always be seen 

 satisfactorily. In the two succeeding sections it is seen to pass back obliquely away from 

 the posterior corner of the slit. Immediately behind it the blood-vessel of the gill-bar 

 emerges from the latter to pass up in the left wall of the pharynx to the aorta. The 

 tubules do not seem to possess an internal opening ; nothing can be made of their 

 histology from my sections, and I cannot say whether or no solenocytes have yet been 



* The appearance of the metapleures in Goldschmidt's fig-ures of A. pelagicus and in mine is somewhat different : 

 ^hia is very probably, however, the effect of a difference of distenaion. I cannot doubt that his specimens and mine 

 liolong to the same species. 



