6oo ON INCREASE IN COMPLEXITY [pt. m 



for producing differential death. An interesting point was that the 

 embryos were first stained with neutral red, and the change in tint 

 of this indicator noted as soon as the killing solutions were poured 

 on them, obviously showing that the differences in killing time of 

 different parts were not entirely due to differences of permeability. 

 Cannon and Huxley were at one time inclined to attribute most of 

 the results obtained by the direct susceptibility method to such 

 differences. But in any case differences of cell-membrane permeability 

 would be included in the variables which might be changing along 

 the physiological gradient. 



The disintegration gradient of the unfertilised lamprey egg was 

 found to be a perfect example of a simple primary gradient, the 

 degeneration beginning at the animal pole and spreading regularly 

 to the vegetative pole. At the 4-cell stage the disintegration begins 

 at the animal tips of the four blastomeres, and passes backwards 

 to meet a slight secondary zone of high susceptibility at the vegetative 

 ends. These changes are seen in Figs. 100 and loi, taken from 

 Hyman's paper. In the 8-, 12- and i6-cell stages, the degeneration 

 begins in the micromeres at the animal pole, then passes on through 

 the macromeres. In later stages the disintegration constantly begins 

 at the animal pole, but in addition isolated cells or groups of cells 

 are to be seen in either animal or vegetal hemisphere which dis- 

 integrate in advance of the region in which they are situated. 

 Hyman supposed that these were taken by the killing solution in 

 the act of cleavage, and were thus more susceptible than their 

 neighbours. The secondary region at the vegetal pole is now dis- 

 appearing for good. The early blastula stages show the usual single 

 gradient, but an important change occurs in the later blastulae, 

 namely, that the spread takes place more rapidly along one surface 

 of the egg than the others. This foreshadows the differentiation of 

 that surface as the dorsal surface. In very late blastulae, there is a 

 small zone of susceptibility near the vegetal pole, which foreshadows 

 the gastrular invagination. The gastrula stages, as shown in the 

 figures, are characterised by disintegration beginning at the anterior 

 end of the embryo and proceeding backwards, but first dorsally 

 and then ventrally to meet the spread from a secondary zone 

 originating around the blastopore. These conditions continue un- 

 changed during the formation of the neural groove and the neural 

 tube, though in the late stages of the latter there is a slight double 



