It2 



NATURE 



[June 13. 1907 



arterial circulation of the kidney does away with the blood 

 current which normally flows a-^'a-j from the region of 

 the tubules, and, this being the case, the venous blood of 

 the renal cardinal meshwork, encountering no resistance, 

 is enabled to penetrate to the tubule plexus, carrying with 

 it the injected diuretics which cause the secretion observed. 

 There is no reason why the secretion should not occur 

 under these abnormal conditions. The tubule epithelium 

 is well supplied with o.\ygcn, the veins arc gorged with 

 impure blood, and in experiments in which at all large 

 secretions were obtained the pressure was artificially raised 

 by forced injection or otherwise. 



It is easy, then, to account for the secretion obtained 

 by the investigators named, and at the same time to believe 

 that the venous blood takes no share in the formation of 

 the kidney secretion under normal conditions. 



It has always surprised me, speaking as an outsider, 

 that physiologists have so readily assumed that they possess 

 in the frog and other animals with " portal " kidneys so 

 many convenient anatomical contrivances in which the 

 glomeruli and the renal tubules are supplied by separate 

 vessels. It is true that the renal cardinal meshwork is 

 continuous with the blood plexus surrounding the tubules, 

 but surely it is very unsafe to assume from this one fact 

 that the venous blood is used by the tubules for secretory 

 purposes. Another equally patent fact, that the similar 

 tubules of mammals employ arterial blood, should suffice 

 t.3 cast doubt on the assumption. And when we recall to 

 mind the statements of Hyrtl iWiener Akad. SB., xlvii., 

 1863) and Vialleton (C. li. Hebtlom. Seances Soc. Biol. 

 Paris, liv., J902), among others, that in those " portal " 

 kidneys in which the vascularisation has been histologically 

 examined, viz. those of the frog and certain sharks, the 

 renal cardinal meshwork is structurally distinguishable 

 from the tubule plexus (the former consisting essentially 

 of large channels putting the post-rena! vein into com- 

 munication with the post-caval, and the latter consisting 

 of capillaries which open into the channels), there is still 

 more reason for supposing that the flow of blood is from 

 the tubule plexus into the renal cardinal meshwork, and 

 not in the contrary direction. The numerous experiments 

 which have been based on the aforesaid assumption have, 

 I should imagine, given rise to incorrect ideas as to the 

 normal functions of the urinary tubules. 



If I needed any additional physiological evidence in sup- 

 port of my contention that the post-renal vein has nothing 

 to do with the vascular plexus of the urinary tubule, i.e. 

 does not supply the kidney for excretory purposes, I find it 

 to hand in the recently issued British Association Report 

 for 1906, York. In a report on " The ' Metabolic Balance 

 Sheet' of the Individual Tissues," p. 427, it is shown to 

 be exceedingly probable, by the relative amounts of oxygen 

 used up by the kidney tissue of a frog and a mammal 

 respectively, that the " renal-portal " vein of the frog bears 

 a very different relation to the kidney tubules as compared 

 with that of the renal vein of mammals — which is the 

 conclusion I am maintaining. It is further stated that 

 " when the same kidney is perfused at different times 

 through the aorta and through the renal-portal system, 

 there is a greater consumption of oxygen in the former 

 case than in the latter (double to treble in four experi- 

 ments)." If we assume what is generally held to be a 

 well-established fact, viz. that the kidney-tubule epithelium 

 plays quite as important a part in kidney secretion as the 

 glomerular epithelium, then it is difficult to understand, on 

 the portal theory of the kidney, whv the quantity of oxygen 

 absorbed by the kidney tubule is totally out of proportion 

 to the work done by it. Obviously the only rational con- 

 clusion to draw is that in the above experiment the o.xygen 

 perfused through the " renal-portal " vein did not come 

 into contact with the tubule. 



To sum up, I think I may say that I have clearly shown 

 that the recent work of Bainbridge, Beddard, and CuUis 

 does not disprove my original contention that the renal 

 cardinal meshwork is, under natural conditions, non- 

 excretory, that, in short, the so-called " renal-portal " vein 

 does not supply the renal tubules, as physiologists 

 commonly assume, and that, in consequence, experiments 

 based on this assumption are liable to give rise to mis- 

 leading ideas. \V. Woodl.ind. 



Mendeltsm and Bionnetry. 



In the striking and suggestive review of Mr. Punnett's 

 work on " Mendelism," in Nature of May 23, the reviewer 

 cites, without naming the author, a view expressed by Mr. 

 A. D. Darbishire (.Manchester Memoirs, 1906) to the effect 

 that " the Mendelian deals with units and the biometrician 

 with masses," and states that this idea, " though 

 plausible, is based on a fallacy," for " the Mendelian 's 

 units are the biometrician 's masses, except when the latter 

 exceeds his limits and includes within his masses itiorc 

 than one such unit." 



I have no doubt that Mr. Darbishire read the review 

 with as much enjoyment as myself, but it seems to me 

 that his statement of the case is dismissed with scarcely 

 sufficient consideration. The reviewer's points, if I under- 

 stand aright, are two: — (1) that Mendel's laws (by which 

 he seems to mean, not merely the law of segregation, but 

 the laws of observed proportions) are really mass-laws 

 and not laws of the individual ; (2) that the biometrician's 

 masses are the masses to which Mendel's laws apply. 

 But surely (i) .Mendel's laws are based on definite con- 

 ceptions as to the germ-cells of the iiiilividual — and that 

 is the important point — and are true of the individual to 

 a degree of approximation which is the higher the greater 

 the number of offspring (quite a high degree in such a 

 case as Mr. Lock's maize). Further, (2) if the " Men- 

 delian's units " were the " biometrician's masses," there 

 should be inheritance of individual variations, within each 

 of two races A and a, for any character to which Mendel's 

 laws applied on crossing those races ; for inheritance of 

 individual variations is what the biometrician has observed 

 for nearly all characters in his masses. 



I indicated the importance of an investigation on this 

 point some time ago (New Phytologist, i., 234) — for it is 

 almost a fundamental question whether a single deter- 

 minant, such as may be assumed to exist for a unit 

 Mendelian character, is or is not capable of variation from 

 individual to individual — but I am not aware that any 

 such investigation has been made. The reviewer's assump- 

 tion may, therefore, be true, but it is unproven, and 

 theories at present in the field (Pearson, Phil. Trans., 

 ic;04 ; Yule, Conference on Genetics, 1906) arc based on 

 the opposite assumption, viz. that heritable individual 

 variations are due to the character concerned being deter- 

 mined by ij allelomorphic couplets, and not by one. If 

 this be true, the " biometrician's masses " are precisely 

 masses to which Mendel's laws, in their simplest form, do 

 not apply. 



The question referred to above, whether a unit 

 Mendelian character exhibits heritable individual vari- 

 ations or no, seems to be one that urgently calls for 

 experimental investigation. G. Udnv Yc'le. 



Mr. YfLE is probably right. The question is this : 

 Is the inheritance which the biometrician always finds 

 within the limits of his masses due to the fact that he 

 is dealing with a large number of Mendelian units, or 

 that he is measuring the intensity of heredity within such 

 a unit? 



If the former, Mr. Yule is right in saying that I 

 criticised the view expressed by Mr. Darbishire unjustly. 

 If the latter, the mass of the biometrician is the unit of 

 the Mendelian. But before we can give an answer to 

 this question we must know, as Mr. Yule points out, 

 whether there is inheritance of fluctuating variations 

 within the limits of a single Mendelian character such 

 as " tall," in peas. If we may argue from stature in man 

 to stature in peas, we should compare the character tall 

 (or normal) in peas to tall (or normal) in man, and dwarf 

 in peas to dwarf in man. We know that there is inherit- 

 ance within the character tall in man, and, if this analogy 

 is legitimate, we should expect to find it so in peas. If 

 it were, the answer to the question whether the view 

 expressed by Mr. Darbishire were right or not would 

 depend on whether we still called the character tall the 

 unit or extended the conception of unit to the smallest 

 heritable variation within the category tall. 



The Reviewer. 



NO. 1963, YOL. 76] 



