348 ORGANIC GROWTH sec. 



moves immediately after its separation iu the same way as 

 the uninjured animal. 



The preceding results lead to tlie inference that a definitely 

 circumscribed nervous system is not absolutely necessary for the 

 exercise of volition. Co-ordinated movement mi^ht also occur 

 if the nerve-cells were uniformly scattered over the whole 

 surface of the body, without a permanent centre. This was 

 tlie original condition, and in a still earlier stage the seat of 

 volition was in the ectoderm cells. The latter is even now 

 the case in some two-layered, free-swimming larvae which 

 exhibit voluntary action, and the presence of special nerve- 

 cells is not indispensable to the activity of e.(j. the Hydra. 

 Quite recently attempts have been made to demonstrate the 

 existence of such cells in Hydra, but it seems to me the 

 evidence is at present insufficient. The fact that each frag- 

 ment of a Hydra, however it is cut up, behaves like an entire 

 animal, and that it grows into an entire animal, that this is 

 true even of a fragment of one of its tentacles, indicates that 

 morphologically expressed centralisation can scarcely exist in 

 this animal. 



That a single will can exist even with a number of brains, 

 is also beautifully shown by the Medusa. I have shown how 

 the eight brains of Aurelia produce single movements : the 

 impulse proceeds always from one of the brains, and passes 

 immediately to the others — sometimes one, sometimes another 

 originating the impulse — but this does not exclude the 

 possibility of all eight acting at the same time, indeed, 

 this is probably necessary for the production of certain 

 movements. 



Now, whether in a given animal there are eight such 

 central points of nervous activity or thousands — as many as 

 there are nerve-cells or ectoderm-cells on the body surface — 

 makes no difference. If we assume that Hydra has no special 

 nerve-cells, then the impulse which produces action may start 



