October 6, 1910] 



NATURE 



451 



found in the neighbourhood of the free chromosomes, a 

 condition which continues until after nuclear division is 

 complete. The absence of potassium, the most abundant 

 basic element in the cytoplasm, would indicate that soaps 

 are not present, and appropriate treatment of such cells, 

 hardened in formaline only, with scarlet red demonstrates 

 that fats, including lecithins, are absent also. This 

 would seem to show that high instead of low surface 

 tension prevails about the nucleus during division. During 

 the " resting " condition of the nucleus this high tension 

 is maintained, for, except in very rare cases, and these 

 of doubtful character, there is no condensation of in- 

 organic salts in the neighbourhood or on the surface of 

 the nuclear membrane. It is also to be noted that the 

 nucleus, with exceptions, the majority of which are found 

 in the Protozoa, is of spherical shape, which also postu- 

 lates that high surface tension obtains either in the cyto- 

 plasmic layer about the nucleus or in the nuclear mem- 

 brane itself. It may also be suggested that high surface 

 tension, and not the physical impermeability of the nuclear 

 membrane, is the reason why the nucleus is, as I have 

 often stated, wholly free from inorganic constituents. 



It does not follow from all tliis that surface tension has 

 nothing to do with cell division. If, as Brailsford Robert- 

 son holds, surface tension is lowered in the plane of 

 division, then the internal streaming movement of the 

 cytoplasm of each half of the cell should be towards that 

 plane, and, in consequence, not separation, but fusion of 

 the two halves would result. The lipoids and soaps 

 would, indeed, spread superficially on the two parts from 

 the equatorial plane towards the two poles, and, accord- 

 ing to the Gibbs-Thomson principle, they would not dis- 

 tribute themselves through the cytoplasm in the plane of 

 division, except as a result of the formation of a septum 

 in that plane. In other words, the septum has first to 

 exist in order to allow the soaps and lipoids to distribute 

 themselves in a streaming movement over its two faces. 

 In Brailsford Robertson's experiment this septum is pro- 

 vided in the thread. If, on the other hand, surface tension 

 is higher about the nucleus in and immediately adjacent 

 to the future plane of division, then constriction of the 

 nucleus in that plane will talve place accompanied or 

 preceded by an internal streaming movement in each half 

 towards its pole, and a consequent traction effect on tlie 

 chromosomes which are thus removed from the equatorial 

 plane. When nuclear division is complete, then a higher 

 surface tension on the cell itself limited to the plane of 

 division would bring about there a separation of the two 

 halves, a consequent condensation on each side of that 

 plane of the substances producing the low tension else- 

 where, and thereby also the formation of the two mem- 

 branes in that plane. 



In support of this explanation of the action of surface 

 tension as a factor in division I have endeavoured to 

 ascertain if, as a result of the Gibbs-Thomson principle, 

 there is a condensation of potassium salts in the cyto- 

 plasm at the poles of a dividing cell, that is, where surface 

 tension, according to my view, is low. The difficulty one 

 meets here is that, in the higher plant forms, cells pre- 

 paring to divide appear to be much less rich in potassium 

 than those in the " resting " stage, and under this con- 

 dition it is not easy to get unambiguous results, while in 

 animal cells potassium may even in the resting cell be 

 very minute in quantity, as, for example, in Vorticella, 

 in which, apart from the contractile stalk, it is limited 

 to one or two minute flecks in the cytoplasm. Instances 

 of potassium-holding cells undergoing division are, how- 

 ever, found in the spermatogonia of higher vertebrates 

 (rabbit, guinea-pig), and in these the potassium is gathered 

 in the form of a minute and thin cap-like layer at each 

 pole of the dividing cell. 



This of itself would appear to show that surface tension 

 is less in the neighbourhood of the poles than at the 

 equator of the dividing cell ; but I am not inclined to 

 regard the fact as conclusive, and a very large number 

 of observations to that end must be made before certainty 

 can be attained. I am, nevertheless, convinced that it is 

 only in this way that we ' can finally determine whether 

 differences of surface tension in dividing cells account, as 

 I believe they do, for all the phenomena of cell division. 

 The difficulties to be encountered in such an investigation 



NO. 



2136, VOL. 84] 



are, as e.xperience has shown me, much greater than are 

 to be overcome in efforts to study surface tension in cells 

 under other conditions, but I am in hopes that what I 

 am now advancing will influence a number of workers to 

 take up research in microchemistry along this line. 



I must now discuss surface tension in nerve cells and 

 nerve fibres. I have stated earlier in this address that 

 I hold that the force concerned in the production of the 

 nerve impulse by the nerve cell is surface tension. The 

 very fact that in the repair of a divided nerve fibre the 

 renewal of the peripheral portion of the axon occurs 

 through a movement — a flowing outward, as it were — of 

 the soft colloid material from the central portion of the 

 divided fibre is, in itself, a strong indication that surface 

 tension is low here and high on the cell body itself. This 

 fact does not stand alone. I pointed out six years ago 

 that potassium salt is abundant along the course of the 

 axon and apparently on its exterior surface, while it is 

 present but in traces in the nerve cell itself. In the latter 

 chlorides also are present only in traces, and therefore 

 sodium, if present, is there in more minute quantities, 

 while haloid chlorine is abundant in the axon. Macdonald 

 has also made observations as to the occurrence of 

 potassium along the course of the axon, and has in the 

 main confirmed mine. We differ only as to mode of the 

 distribution of the element in the axon, and the manner 

 in which it is held in the substance of the latter ; but, 

 whichever of the two views may be correct, it does not 

 affect what I am now advancing. E.xtensive condensation 

 or adsorption of potassium salts in or along the course 

 of the axon, while the nerve cell itself is very largely free 

 from them, can have but one explanation on the basis of 

 the Gibbs-Thomson principle, and that explanation is that 

 surface tension on the nerve cell itself must be high while 

 it is low on or in its axon. 



The conclusions that follow from this are not far to 

 seek. We know that an electrical displacement or dis- 

 turbance of ever so slight a character occurring at a 

 point on the sijrface of a drop lowers correspondingly the 

 surface tension at that point. What a nerve impulse 

 fundamentally involves we are not certain, but w-e do 

 know that it is always accompanied by, if not constituted 

 of, a change of electrical potential, which is as rapidly 

 transmitted as is the impulse. When this change of 

 potential is transmitted along an axon through its synaptic 

 terminals to another cell, the surface tension of the latter 

 must be lowered to a degree corresponding to the magni- 

 tude of the electrical disturbance produced, and, in con- 

 sequence, a slight displacement of the potassium ions 

 would occur at each point in succession along the course 

 of its axon. This displacement of the ions as it proceeded 

 would produce a change of electrical potential, and thus 

 account for the current of action. The displacement _ of 

 the ions in the axon would last as long as the alteration 

 of surface tension which gave rise to it, and this would 

 comprehend not more than a very minute fraction of a 

 second. Consequently, many such variations in the surface 

 tension of the body of the nerve cell would occur in a 

 second; and, as the physical change concerned_ would 

 involve only the very surface layer of the cell, a minimum 

 of fatigue would result in the cell, while little or none 

 would develop in the axon. 



It mav be pointed out that in medullated nerve fibres 

 the lipoid-holding sheath, in close contact as it is with the 

 axon, must of necessity maintain on the course of the 

 latter a surface tension low as compared with that on the 

 nerve cell itself, which, as the synaptic relations of other 

 nerve cells with it postulate, is not closely invested with 

 an enveloping membrane. In non-medullated nerve fibres 

 the simple enveloping sheath may function in the same 

 manner, and probably, if it is not rich in lipoid material, 

 in a less marked degree. 



What further is involved in all this, what other con- 

 clusions follow from these observations, I must leave 

 unexplained. It suffices that I have indicated the main 

 points of the subject, the philosophical significance of 

 which will appear to those who will pursue it beyond the 

 point where I leave it. 



In bringing this address to a close, I am well aware of 

 the fact that my treatment of the subjects discussed has 

 not been as adequate as their character would warrant. 



