226 



PERCY SLADEN TEUST EXPEDITION. 



B 



The accompanying figures (text-figs. 1 & 2) show how the various ridges (1-6) run 



into one another at the fore and hind margins of the bar — 

 being in reality no more tlian a single ridge or fold in the 

 vertical plane, thrown into several folds in the horizontal plane. 

 Goldschmidt's fig. 18 shows it equally well, if we eliminate the 

 fiction of the continuous sagittal furrow. This furrow is 

 merely the result of the folding, and has no more morpho- 

 logical significance than the lateral ones : somewhat further 

 back than the 14tli gill-slit, indeed, it loses its "sagittal" 

 position, the ridge numbered 6 in the figures being no longer 

 jjresent, and a ridge (3) instead of a furrow occupying the 

 " sagittal " plane (text-fig. 2, B). 



What I have said of the folds hitherto applies to A. valdioice : 

 in A. 2^^l('U^ctis they are of similar character, but even more 

 complex, projecting right up into the " pars nutritoria " between 

 the lateral folds of tlie pliaryngeal wall. 



Undoubtedly the gill-slits do show a more or less symmetrical 

 disposition about the middle line, particularly marked in their 

 musculature, but — theory apart — I do not think it is possible 

 to see in them the essentially " paarige Aufbau " claimed by 

 Goldschmidt. 



The typical number of gill-slits appears to be, for A. loela- 

 gicus, 27, attained simultaneously with or a little after the full 

 number of myotomes; for A. valdivice, 34, attained when the 

 animal is about 8'5 mm. long. Further details are given in 

 the table on p. 218. 

 As to the apparent eumetam ery of the slits in A. 2)elagicus I can fully corroborate 

 Goldschmidt: the 27th slit coincides with great regularity with the ventral end of the 

 septum dividing 27th and 28th myotomes ; I only observed one exception. In younger 

 specimens, as he did, I found that the correspondence was not exact ; thus in an animal 

 with 16 slits, the last corresponded with the line of division between 19th and 20th 

 myotomes. He suggests, to account for this lack of correspondence, that the most anterior 

 gill-slits may be formed late. In view of their known sequence of development in the 

 Branchiostoma larva, where the first appears long before the others, and the rest in 

 regular order from before backwards (Lankester & Willey, 1890), this seems hardly 

 probable. It would seem more likely that the segmentation of the gill-slits is after all 

 independent of that of the myotomes, especially in view of their non-correspondence 

 in A. valdivice. In the latter Goldschmidt regards a vai-iable number of the anterior 

 slits (determined by the excess of the number of slits over that of the corresponding 

 myotomes) as pre-somital, and supposes the splanchnocoele in relation to them to represent 

 the ventral mesoderm of so many formerly-existing somites, the dorsal parts of which 

 are now incorporated into the first myotome-pair and lateral rostral canals. Thus he 

 holds that the whole region of the pharynx lying between the 5th-8th gill-slit and the 



Fig. 2. — A. Endodermal face 

 of 14th gill-bar, with out- 

 lines of the gill-slits in 

 front and behind. (Kecon- 

 struction.) B. Similar re- 

 construction of endodermal 

 face of a gill-bar further 

 back. 



