August 9, 1877] 



NA TURE 



291 



in the one case to sting away the enemy, and in the other 

 to capture the prey. Hence I think that natural selection 

 would probably tend to convert lines of discharge in pro- 

 miscuous directions, into lines of discharge in definite 

 directions — thus developing the function of localisation. 

 At first, no doubt, this function would be performed only 

 in a general and tentative manner (as, indeed, I have 

 observed in the case of ^//;r//(0 ; but gradually by the 

 combined action and mutual reaction of use and survival 

 of the fittest, this function would come to be performed 

 with ever-increasing precision.' 



This, then, t conceive to be an important step in the 

 evolution of nen'Ous systems — foreshadowing as it does 

 the principle of coordination among muscular move- 

 ments, which in all the higher animals is effected by 

 reflex mechanisms precisely resembling, as to their func- 

 tion, the primitive reflex mechanism we are considering. 

 But now another point of interest arises. As Spencer's 

 theory supposes a line of discharge to become more and 

 more definite by use, if, for the maintenance of any par- 

 ticular function such as the one we are considering, a 

 certain line of discharge habitually serves as a line of 

 communication between two points of the animal tissues ; 

 it follows that this line will offer less resistance to the 

 passage of a stimulus bet-veen these two points than would 

 any other line in the organism. Consequently, so long as 

 such a line remains intact, so long we shou'.d expect what 

 we have seen to be the case, viz., that little or no vicarious 

 action takes place between it and other lines. But let 

 this line be severed, and let there be a number of closely 

 adjacent lines, as there must be in this particular instance, 

 and should we not expect, both from Spencer's theory and 



Fig. 7. — Tia?-o/-sis imlu-nns. ilis'atly enlarged. 



from our knowledge oi Aitreiia, that at some such grade 

 of nervous evolution as Tiai'opsis presents, the stimulus 

 should be able to escape from the severed to the unsevered 

 lines ? And this I find to be the case. For if a small cut 

 be introduced between the base of the polypite and the 

 .•■eat of injury in the bell, the polypite is no longer able to 

 localise the seat of injury, although it still continues to 

 pfrceive, so to speak, that injury is being applied some- 

 luliere. For instance, if a short cut be introduced as here 

 represented at l> c, and you prick the bell anywhere below 

 the cut, as at (/, the pol>pite, instead of immediately 

 applying its extremity to the exact spot that is being 

 stimulated, now actively dodges about first to one part 

 and then to another part of the bell, as if seeking in vain 

 f<ir the offending body, which, however, it cannot succeed 

 in finding. Now I explain this marked change in the 

 behaviour of the polypite by supposing that the wave of 

 stimulation in this case runs along the habitual line of 

 discharge till it reaches the cut ; and that being there no 

 longer able to pursue this habitual line of least resistance, 

 the wave of stimulation escapes into the adjacent lines, and 

 so spreads all over the bell. Hence a number of conflict- 

 ing messages are simultaneously delivered to the polypite, 

 wtiich therefore executes the random movements I have 



* It may be here observed that Mr. Spencer, in his theory of nervo-geneti=, 

 expre'.sly supplements his hypothesis as to the direct influence of u-e, with 

 that as to the indirect influence of natural selection. (See " Biology/" 

 § i6,.)-G. J. K. 



described — each of these movements being presumably 

 determined by the relative degree in which now one line 

 and now another takes part in conveying the scattered 

 stimulus. 



And now for another expectation to be realised. We 

 should expect that the higher degree of specialisation 

 which in these lines of discharge prevents vicarious action 

 so long as the lines are undivided, should have the effect 

 of rendering such vicarious action as we have seen to 

 ensue when the lines are divided, less easy than it is in 

 Aurelia, where the specialisation of the lines being less 

 pronounced, vicarious action among them is presumably 

 more habitual. And such I find to be the case ; for while 

 in Aurelia, as we have seen, stimulus-waves continue to 

 zig-zag round and round the ends of almost any number 

 of overlapping cuts, in Tiaropsis ttvo or three such cuts 

 are sufficient to destroy, not only the localising, but also 

 the random movements of the polypite — the latter then 

 remaining passive, because the stimulus-waves are wholly 

 blocked. 



And lastly, before leaving the case of Tiaropsis indicans, 

 I should like to mention the noteworthy fact, that although 

 the polypite is able to perform the intricate ganglionic 

 finiction of localising any seat of stimulation in the bell, 

 no signs of ganglionic structure can be detected with the 

 microscope. Moreover, a portion of any size that is re- 

 moved from the polypite continues to perform the localis- 

 ing function in just the same way as does the entire organ. 

 In other words, this localising function, which is so very 

 efficiently performed by the polypite of this Medusa, and 

 which, if anything resembling it occurred in the higher 

 animals, would certainly have definite ganglia for its 

 structural correlative, is here shared equally by all parts of 

 the exceedingly tenuous excitable tissue that formsthe 

 outer surface of the organ. The case of the incipient 

 ganglia of the polypite thus resembles that of the incipient 

 nerves of the bell in this respect— that in both cases 

 obvious signs of characteristic function are displayed 

 before any corresponding signs of structure can be dis- 

 tinguished. Nerve-cells, therefore, no less than nerve- 

 libres, are thus shown to have their first beginnings in 

 d flerentiations of protoplasmic substance which are too 

 refined for the microscope to analyse. 



There is one other species of Medusa about which I 

 should like to say a very few words, because it presents a 

 still higher grade of nervous evolution than Tiaropsis. This 

 is Sarsia(Y\g. 8), a Medusa in which the lines of discharge 

 have in some places become so far differentiated as to 

 admit of being actually seen, and are therefore entitled to 

 be called nerves. All round the margin, and likewise along 

 the course of the radial tubes, these, the earliest visible 

 nerve-fibres in the animal kingdom, may be traced. And 

 as we might anticipate, the advance of stiucture which is 

 implied by an invisible " line of discharge" becoming a 

 visi'ole nerve-fibre, entails a corresponding advance of 

 function. In the first place, the rate at which a stimulus 

 travels seems to be much greater along these fully-evolved 

 nerve-fibres than it is in the more rudimentary nerves or 

 lines of discharge in Aurelia. In the next place, this 

 greater differentiation of nerve-tissue renders the nervous 

 connection between any two parts of the organism much 

 more definite, and therefore vicarious action less promis- 

 cuous, than we have seen it to be in the other jelly-fishes ; 

 so that, for instance, a tentacular wave in this species may 

 be blocked by a single short cut through the margin of 

 the bell. Lastly, it is in this species that I was first able 

 to perceive any unequivocal evidence of co-ordination 

 among the marginal ganglia. In all the other species of 

 Medusse the marginal ganglia appear to act independ- 

 ently of one another; but in this species, where the mar- 

 ginal ganglia are first seen to be united by a visible 

 nerve-fibre, they always act in concert. So much, indeed, 

 is this the case, that the animal is able to steer itself m 

 any required direction, as proved by the experiment which 



