424 PROCEEDINGS OF THE AMERICAN ACADEMY. 



just defined. Hesse ('98^, p. 457) states that in Branchiostoma lan- 

 ceolatum the eye-cups are most abundant anteriorly and diminish 

 in numbers posteriorly, till in the tail there may be not more than 

 one cup to a segment. But this description, as Boeke ( : 02, p. 352) 

 and Joseph (:04, p. 18) have noted, is somewhat defective. In five 

 specimens of B. lanceolatum from Naples that I have examined, the 

 distribution was essentially like that in B. caribbaeum, in that, in ad- 

 dition to the considerably increased number of cups anteriorly, there 

 was also an increase in the number in the tail region. This confirms 

 Joseph's statement (:04, p. 18) for this species and agrees with the 

 discovery of Boeke (:02, p. 352), that in young pelagic individuals of 

 B. lanceolatum there are to be seen two groups of eye-cups, one anterior 

 and the other posterior, corresponding to the two concentrations men- 

 tioned. These two groups presumably unite later to form one series. 

 The general plans of distribution of the cups in the two species, then, 

 undoubtedly agree, and, since these plans of distribution correspond 

 to the different degrees of sensitiveness to light for the different parts 

 of the body in B. caribbaeum, I believe that the eye-cups described by 

 Hesse, and not the skin, are the light-receptive organs. 



In Branchiostoma caribbaeum, as in B. lanceolatum according to 

 Hesse ("98'', p. 458) and Boeke (:02, p. 351), the ventral eye-cups, as 

 well as those of the right side, point in the main ventrally, while those 

 of the left side point mostly dorsally. Hesse states further that in 

 B. lanceolatum the cups of the two sides tend toward the right, and he 

 suspected that this might be correlated with a possible habit of resting 

 on a particular side. But in testing this hypothesis Hesse ('98'', 

 p. 459) found that the animals rested about as frequently on one side as 

 on the other, and he therefore abandoned it. In B. caribbaeum I could 

 not see that the cups were directed more toward the right than toward 

 the left, but it was apparent that the majority pointed ventrally. This 

 position seemed to me entirely consistent with the habits of this species, 

 for it naturally lies in the sand with the ventral side obliquely xqiper- 

 most, the majority of eyes being thus directed toward the most prob- 

 able source for light. However, individuals that were in a glass dish 

 without sand were, so far as I could see, equally sensitive to light fall- 

 ing on them in any direction. 



If the Hght-receptive organs in amphioxus are the eye-cups of the 

 nerve-tube, any part of the animal containing these organs might be 

 expected to retain its sensitiveness to light. Nagel (94"^, p. 811 ; '96, 

 p. 79), after cutting these animals in two transversely, found that both 

 halves still reacted promptly to light, but less energetically than the 

 whole animal did. Krause ('97, p. 514) declared that after halving 



