234 PEllCY SLADEN TRUST EXPEDITION. 



Corresponding with the 32nd segment is seen the short " liver " caecum. There appear 

 to be 15 laterally situated gill-slits on the left, with indications of two more behind 

 them, which may he median in position. The precise number of the right gill-slits, 

 which resemble the left in appearance, I am unable to determine. The first slit is very 

 small and undivided ; it is followed by 3 in which the ventral ends of the tongue-bars 

 ai-e not fully formed ; tlie 4th slit is the largest of the series, and — with the five 

 succeeding — is completely divided by its tongue-bar into two equal halves ; the slits behind 

 the 10th are small and incompletely divided. 



In place of the pre-oral organ of Amphioxides we find a buccal hood into both cheeks 

 of which the ventral ends of the myotomes extend, not being crowded together and 

 carried back as in Amphioxides; it is not yet fully formed, the left cheek being 

 considerably shorter dorso-ventrally than the right, and the opening still facing towards the 

 left. Round the posterior end of the opening the buccal skeleton has begun to form : 

 its lower, definitive right arm already shows 6 cirrlii, its upper 4, the most posterior 

 being the longest in both cases, while there is also a long median cirrhus ; the cirrhi are 

 perfectly smooth, as in Asymmetron, and have not the sensory thickenings found in 

 Branchiostoma and Meteropleuron. Behind the buccal skeleton the velum can be dimly 

 seen, and appears to be shaped like a V with its apex posterior. 



No gonads could be seen in this specimen. 



Internal Anatomy. 



This M as studied in sections of the la,rger specimen (PI. 15. figs. 13-18). The latter was 

 8'6 mm. when first measm-ed; the greatest depth of the sections is '66 mm., giving 

 13 as the ratio of length to depth ; since shrinkage may safely be assumed to have 

 occurred, this figure is certainly too large. Neither can the ratio of total to jjost-anal 

 length — apparently about 6'5— be relied upon, since it was obtained by reconstruction, 

 and was, unfortunately, not observed in the uncut specimen. As in A. 'pelagicus, 

 the number of myotomes was 67, the anus corresponding with the 53rd, and the origin 

 of the dorsal fin-membrane with the 25th or thereabouts; the "liver" csecum lay 

 opposite to the 32nd myotome, the atriopore to the 42nd and 43rd. 



Both metamorphosed animals almost certainly represent the same species, and are in 

 A ery much the same stage of development, though tliey do not agree absolutely in their 

 myotome formulae and number of gill-slits. 



The cavities of the rostrum show almost identical relations with those of 

 A. pelagicas, the only difference affecting the ventral rostral cavity, the anterior part 

 of which is occluded and forms a solid lamella of " skeletal " tissue similar to that 

 representing the " dorsal rostral canal " of Goldsehmidt. Behind the level of the brain 

 a cavity appears in this lamella (immediately ventral to the consi^icuous ventral rostral 

 canal), but it breaks up again into lymph-canals roiurd the anterior end of the invagination 

 of the buccal hood ; one of these canals passes back for a short distance in the left cheek 

 of the hood, the rest, lying to the right of the invagination, re-unite to form a spacious 

 cavity which is directly continuous with the right division of the stomocoele. No trace 



