Decembee 2, 1904.1 



SCIENCE. 



743 



not improbable that the anterior piece also, 

 if kept alive, would regenerate a head on 

 its posterior cut surface. 



I have obtained a similar result with the 

 leg of one of the salamanders, Spelerpes 

 ruber. If the foot is first cut off and then 

 the skin of the foreleg is loosened, so that 

 a piece of the internal parts can be cut out, 

 and if this piece is turned round, and 

 grafted in the pocket made by the loosened 

 skin, there regenerates from the free end 

 a new foot (and not a salamander). 



A few other examples might be given, 

 but these will suffice to illustrate the main 

 points. The new part has in these cases 

 only one possibility, and the same struc- 

 ture regenerates from the anterior and 

 from the posterior cut surface, regardless 

 of the 'polarity' of the old part. 



The interpretation of these facts is, I be- 

 lieve, not difficult in the light of certain 

 other results that are now known. Only 

 one kind of structure develops because the 

 group of new cells that appears over the 

 cut end is such that out of it only one kind 

 of organ can be formed. This may sound 

 paradoxical, but my meaning will, I hope, 

 be clearer when we have considered the 

 next category, when an alternative exists. 



2. REGENERATION WHEN AN ALTERNATIVE 

 EXISTS. 



The worm Lumbriculus furnishes the 

 most striking instance of this sort. If the 

 worm is cut in two at almost any level a 

 head regenerates from the anterior end 

 and a tail from the posterior end. Since 

 the two cut ends are identical, each must 

 have both potentialities. Nevertheless, a 

 head forms at one end and a tail at the 

 other. Planarians give the same results. 

 Pieces of Hydra also behave in the same 

 way. Another hydroid, Tuhularia, often 

 produces a head (hydranth) at one (apic- 

 al) end and a stolon at the other; but also 

 quite frequently produces a head at both 



ends. It has been shown, in fact, in all 

 these forms, except Lumbriculus, that a 

 head may regenerate on the posterior end 

 under certain conditions. 



Thus, if the apical end of Tuhularia be 

 tied or stuck into the sand, a head develops 

 on the basal end. If a piece is sharply 

 bent the same result happens. 



If very short cross-pieces of Planaria 

 m.aculata be cut out, a head often develops 

 on both ends. If two pieces of hydra are 

 grafted together by their anterior ends, 

 one piece being longer than the other, the 

 'polarity' of the shorter piece will be re- 

 versed, and a single hydra regenerate. 



Several recent writers have attempted to 

 account for these eases of reversal on the 

 old Bonnet-Sachs hypothesis that form- 

 ative stuffs migrate in definite directions. 

 Loeb, for example, has tried in a recent 

 paper on Tuhularia to rehabilitate this 

 view, but, I believe, without success. Dur- 

 ing the past summer I have carried out a 

 large number of experiments on Tuhu- 

 laria* which I think show that the as- 

 sumption of the migration of stuffs in 

 definite directions is not needed to explain 

 the results, and there is in reality nothing 

 in the experiments to support such an idea. 



It would take me too long to go into the 

 details necessary to substantiate this state- 

 ment; but in a forthcoming paper I shall 

 hope to discuss the question more fully. 

 It must suffice here to state that, in my 

 opinion, the development of a head on the 

 apical or on the basal end of Tuhularia may 

 be explained if we assume that the amount 

 of nutritive substances present at a given 

 moment represents one of the internal 

 factors and the stimulus of the sea water 

 on the more responsive free end (which is 

 always that nearer to the old head) repre- 

 sents the external factor that calls forth 

 the regeneration of the hydranths. , 



The third category applies only to plants. 



* In collaboration with N. M. Stevens. 



