450 



NATURE 



[OCTOUER 6, 1 910 



fused with potassium salts condensation products like 

 diphenol are produced, but when sodium salts are used 

 the products are dioxybenzol and phloroglucin (Barth). 

 It is, therefore, not improbable that potassium, along 

 with those properties which come from its ionic mobility, 

 has a special value in the metabolism of the dim bands 

 of striated muscle fibre and in the condensation synthesis 

 which characterise the chromatophors of protophyta 

 (Spirogyra, Zygnema). 



With the use of this method of determining differences 

 in surface tension in cells it is possible, in some cases 

 at least, to ascertain whether this force plays a part in 

 both secretion and excretion, and evidence in favour of 

 this view can be found in the pancreatic cells of the 

 rabbit, guinea-pig, and in the renal cells of the frog. In 

 the pancreatic cells there is an extraordinary condensa- 

 tion of potassium salts in the cytoplasm of each cell 

 adjacent to the lumen of the tubule, and during all the 

 phases of activity — except, it would appear, that of the 

 co-called " resting stage " — potassium salts occur in, and 

 are wholly confined to, this part of each cell. It is 

 difficult to say whether they pass into the lumen with 

 the secretion and their place taken by more from the 

 blood-stream and lymph, but the important point is that 

 the condensation of potassium salts immediately adjacent 

 to the lumen seems to indicate a lessened surface tension 

 on the lumen surface of the cell. 



According to Stoklasa,' the pancreas of the pig is much 

 richer in potassium than in sodium, the dried material 

 containing 2oq per cent, of potassium and 0-28 per cent. 

 of sodium, while the values for the dried material of ox 

 muscle are, as he determined them, 1-82 and 0-26 per cent, 

 respectively. It is significant that in the pancreas this 

 large amount of potassium should be localised as 

 described. 



In the renal cells of vertebrates there is usually a con- 

 siderable amount of potassium salts distributed through- 

 out the cytoplasm. These cells are always active in the 

 elimination of the element from the blood, and it is in 

 consequence not possible to determine whether there are 

 differences in surface tension in them. Under certain 

 conditions, howevej", these can be demonstrated. In the 

 frogs which have been kept in the laboratory tanks 

 througliout the winter, and in the blood of which the 

 inorganic salts have been, because of the long period of 

 inanition, reduced to almost hypotonic proportions, the 

 renal cells are very largely free from potassium. When 

 it is present it is usually diffused throughout the cyto- 

 plasm. If now a few cubic centimetres of a decinormal 

 solution of potassium chloride be injected into the dorsal 

 lymph sacs of one of these frogs, and after twenty minutes 

 the animal is killed, appropriate treatment, with the 

 cobalt reagent, of a thin section of the fresh kidney made 

 by the carbon dioxide freezing method, reveals in the 

 cells of certain of the tubules a condensation of potassium 

 salts in the cytoplasm immediately adjacent to the wall 

 of the lumen. There is also a very slight diffuse reaction 

 throughout the remainder of the cytoplasm, except in 

 that part immediately adjacent to the externa! boundary 

 of the tubule. In these cells the potassium injected into 

 the lymph circulation is being excreted, and the condensa- 

 tion of the element at or near the surface of the lumen is 

 evidence that there the tension is less than at the other 

 extremity of the cell. 



These facts are in their significance in line with some 

 observations that I have made on the absorption of soluble 

 salts by the intestinal mucosa in the guinea-pig. When 

 the " peptonate '* of iron was administered in the food of 

 the animal it was not unusual to find that in the epithelial 

 cells of the villi the iron salt was distributed through the 

 cytoplasm, but its concentration, as a rule, was greatest 

 in the cytoplasm adjacent to the inner surface of the 

 cell, from which it diffused into the underlying tissue. 

 Here also, inferentially, surface tension is lower than 

 elsewhere in the cell. 



It would perhaps be unwise to form final conclusions 

 at this stage in the progress of the investigation of the 

 subject, but the results so far gained tempt one to adopt 

 as a working hypothesis thai in the secretin^; or the 

 excreting cell lower surface tension exists at its secreting 

 1 ."Stokl.isa gave the valu-^ in K2O and NaoO. 



NO. 2136, VOL. 84] 



or excreting surface than at any other point on the cell 

 surface. How this low surface tension is caused or main- 

 tained it is impossible to say, but, whatever the solution 

 of the question may be, it is important to note that we 

 must postulate the participation of this force in renal 

 excretion in order to explain the formation of urines of 

 high concentration. These have a high osmotic pressure, 

 as measured by the depression of the freezing point, while 

 the osmotic pressure of the blood plasma determined in 

 the same way is low. On the principle of osmosis alone, 

 as it is currently understood, this result is inexplicable, 

 for the kinetic energy, as required in the gas theory of 

 solutions, should not be greater, though it might be less, 

 in the urine than in the blood. It is manifest that in 

 the formation of concentrated urines energy is expended. 

 We know also from the investigations of Barcroft and 

 Brodie that the kidney during diuresis absorbs much more 

 oxygen per gram weight than the body generally, and 

 that, assuming it is used in the combustion of a proteid, 

 a very large amount of energy is set free, very much 

 more, indeed, than is necessary. It has also been 

 observed that a portion of the energy set free is found in 

 a higher temperature in the excretion than obtains in the 

 blood itself circulating through the kidney. This large 

 expenditure of energy is, probably, a result of the physio- 

 logical adaptation of the principle of the " factor of 

 safety," which, as Meltzer has pointed out, occurs in other 

 organs of the body. 



In cell and nuclear division surface tension operates 

 as a force, the action of which cannot be completely 

 understood until we know more of the part played by the 

 centrosomes and centrosphere. That this force takes part 

 in cell reproduction has already been suggested by Brails- 

 ford Robertson. He has devised an ingenious experiment 

 to illustrate its action. If a thread moistened with a 

 solution of a base is laid across a drop of oil in which 

 is dissolved some free fatty acid, the drop divides along 

 tlie line of the thread. When the latter is moistened with 

 soap the drop divides in the same way and in the same 

 plane. The soap formed in one case and present in the 

 other, it is explained, lowers the surface tension in the 

 equatorial plane of the drop, and this diminution results 

 in streaming movement away from that plane which bring 

 about the division. He suggests that in cell division 

 there is a liberation of soaps in the plane of division which 

 set up streaming movements from that plane towards the 

 poles, and terminating in the division of the cytoplasm 

 of the cell. 



I have observed in the cells of Zygnema about to divide 

 a remarkable condensation of potassium in the plane of 

 division. In the " resting " cell of this Alga the potass- 

 iu.Ti is, as a rule, more abundant in the cytoplasm near 

 the transverse walls of the thread, and only traces of the 

 clement are to be found along the line of future division 

 of the cell. But immediately after division has talcen 

 place the potassium is concentrated in the plane of 

 division. Ttiis would seem to indicate that surface tension 

 in the plane of division is, as postulated by the deduction 

 from the Gibbs-Thomson principle, lower than it is on 

 the longitudinal surface, and lower, especially, than it is 

 on the previously formed transverse septa of the thread. 



One must not, however, draw from this the conclusion 

 that in all dividing cells surface' tension is lower in the 

 plane of division than it is elsewhere on the surface of the 

 dividing structure. All that it means is that in the 

 dividing cell of Zygnema the condition already exists along 

 the plane of division, which subsequently makes for low 

 surface tension in the cell membrane immediately adjacent 

 to each transverse septum in the confervoid thread. If 

 the evidence of low surface tension vanished immediately 

 after division was complete, then it might be held that it 

 determined the division. .As it is, the low surface tension 

 in this case is the result, and not the cause, of the 

 division. 



This conclusion is corroborated by the results of 

 observations on the cells of the ovules of Lilium and 

 Tulipa. The potassium salts in these are found condensed 

 in minute masses throughout the cytoplasm. When 

 division is about to begin the salts are shifted to the 

 prripheral zone of the cytoplasm, and when the nuclear 

 membrane disappears not a trace of potassium is now 



