GIBSON— CEPHALOCHORDA : " AMPHIOXIDES." 



225 



A more important point upon which I differ has to do with the structure of the endo- 

 dermal portions of the burs. Gohlschmidt derives important support for his view, that 

 the single series of gill-shts of Amphioxides is morphologically median and primitive, 

 from the bilaterally symmetrical structure which he professes to see in the bars. He 

 elucidates this bilateral structure by means of a diagrammatic schema (p. 35), and writes : 

 " Das Schema zeigt uns schUesslich noch, dass der Darmanteil des Kiemenbogens durch 

 eine tiefe mediane Furche in zwei symmetrischen Halften geschieden wird, so also ein 

 paariger Auf l)au des unpaaren Gebildes angedeutet ist." I must confess that I cannot 

 find his figui'es made from actual preparations nearly so convincing on this point as the 

 schema ; in the excellent drawings from models — for instance, figs. 12, 13 — the right 

 appears to predominate considerably over the left, even allowing for the partial conceal- 

 ment of the latter. The two symmetrical halves of each bar, according to Goldschmidt, 

 are represented by projections (" Erhebungen ") of the endoderm separated by a deep 

 sagittal furrow — each projection further subdivided by a furrow cutting into it from m 

 front. But he does not note that the "sagittal" furrow is itself interrupted at the anterior 

 margin of each bar by the junction of its two boxinding ridges, so that we might with 

 equal justice speak of a median " Erhebung" into which the furrow cuts from behind. 

 The cessation of this furrow, anteriorly, is not so obvious as that of the lateral ones, 

 posteriorly, for the reason that the tongue of tissue in which it ends is directed ventrally 

 (thus it is somewhat hidden from view in his fig. 13 a). This tongue of tissue is in fact, 

 the " medial Einbuchtung " which Goldschmidt describes at the bottota of p. 37, but by 

 a curious mistake places at the hinder margin of the gill-bar. In reality there is no 

 median projection at the posterior, but only at the anterior margin of the bar, as is 



Fig. 1.— A series (interrupted) of transverse sections through the 14th gill-bar of A. vaUivue. The first section 

 (top, left-hand side) is the most anterior, and the last (bottom, right-hand side) the most posterior. 



obvious from his fig. 18. His figs. 19 and 20 would be correct representations of gill- 

 bars seen from the rkjht, not from the left side. 



