508 



SCIENCE 



[N. S. Vol. XLIV. No. 1136 



tion of water as the result of a change in 

 those surfaces of the absorbent cells which are 

 exposed to the inconstant intra-embryonic en- 

 vironment. If this change involves a weaken- 

 ing of the face of the neural plate that becomes 

 convex, the curvature that leads to the forma- 

 tion of a tube would be accounted for. 9 



Accordingly then, the absorption of water is 

 not the cause of folding, but a symptom of that 

 cause. If this interpretation is correct, the 

 water content of the cells at any given level in 

 the early stages of involution can not be uni- 

 form. In fact the theory demands that the 

 marginal cells of the neural plate, the first, it 

 will be recalled, to undergo a change of shape, 

 shall have a higher water-content than the 

 cells in the middle of the plate which only 

 assume the wedge-shape during the last stages 

 of involution. 



For the decision of this crucial question, no 

 direct method is as yet available. However, it 

 is possible to secure evidence indirectly which 

 seems to me convincing. 



If the eggs of the starfish are placed in 

 hypotonic sea-water, and given an opportunity 

 to absorb more water than they normally con- 

 tain, they at once increase in volume, and 

 their nuclei, easy to deal with on account of 

 their spherical shape, also enlarge. The facts 

 on which this statement is based are given in 

 Table Y. 



TABLE V 



Asterias Eggs in Various Concentrations of Sea- 

 water 



o For the relation between this view and the 

 Rhumbler Surface-Tension Hypothesis, as well as 

 for a criticism of the latter, see Glaser, loc. cit., 

 pp. 536-548. 



Before applying this information to the 

 problem in hand, I had first of all to deter- 

 mine whether these facts held for the nervous 

 system, and especially whether measurable 

 differences could be demonstrated in those re- 

 gions known to have contained during life, 

 different proportions of water. 10 



TABLE VI 



Relative Water Contents of Embryonic Cords and 

 Brains 

 Embryonic Cords Embryonic Brains 



Amblystoma 125 > 125 by 2.2 per cent 



Sana 139 > 135 by 1.9 per cent. 



Sana 192 > 188 by 2.3 per cent. 



Relative Sizes of Nuclei 

 Nervous System of Cryptobranchus Embryos 

 Stage I Stage II Stage III 



Cord Brain Ratio Cord Brain Ratio Cord Brain Ratio 



109 109 1:1.2 126 114 1:1.1 125 113 1:1.3 



119 114 1:1.2 107 108 1:1.1 143 123 1:1.2 



120 125 1:1.1 127 129 1:1.2 133 113 1:1.3 



121 119 1:1.1 115 131 1:1.3 



Control 36-hour Chick 

 End of Cord Forebrain Ratio 



110 124 1:1.4 



Relative Sizes of Nuclei in Center and at Edges of 

 Neural Plate in Cryptobranchus 



during Folding 

 Number and Positions of Nuclei 

 Central Lateral Ratio 



110 115 1:1.2 



112 120 1:1.2 



112 111 1:1.1 



135 122 1:1.2 



In both Amblystoma punctatum and Rana 

 pipiens (Table VI.), a comparison of the ante- 

 rior and posterior ends of the embryonic nerv- 

 ous systems, indicates a higher water-content 

 in the larval brain than in the cord. Since 

 these results are consistent, and, in sense, agree 

 with corresponding differences found by 

 Donaldson (loc. cit.) for the adult nervous 

 system of Sana pipiens, I feel fairly certain 

 of the essential correctness of my values, and 



if That the embryonic brain has a higher water 

 content than the cord is indicated by the figures 

 which I published in Science, N. S., Vol. XXXIX., 

 pp. 730-731, in 1914. The evidence there pre- 

 sented was meager and, unfortunately, I overlooked 

 some arithmetical errors. Recalculation has made 

 no essential difference in the results, however, and 

 further evidence now shows them to have been es- 

 sentially correct. 



