2 74 



NATURE 



[March 2, 1922 



there are very many such small areas, we see that 

 there will be very many species which occupy small 

 areas. 



We can obtain, in fact. Dr. WilUs's (number of 

 species) against (size of area) curve simply by assum- 

 ing (i) that in any very large area the distribution 

 of different kinds of environment is random, and 

 (2) that organisms are adapted to their environment. 



I therefore come to conclusions exactly opposite to 

 those of Dr. Willis, for I think we have in his curves 

 direct evidence that : — 



(i) Evolution has proceeded almost entirely by 

 natural selection adapting a species to the limitations 

 of its environment. 



(2) Animals are so closely adapted to their con- 

 ditions of existence that it is impossible to conceive 

 of evolution proceeding by the large mutations 

 suggested by Dr. Willis. 



Space prevents me considering here Dr. WiUis's 

 (number of genera) against (number of species) curves, 

 but these also are susceptible of a similar interpreta- 

 tion in terms of natural selection. 



C. F. A. Pantin. 



Christ's College, Cambridge, February i6. 



We find it very difificult to follow the hypotheses 

 made by Mr. Pantin in his interesting letter, and 

 cannot agree that they accord with reality. We can- 

 not see how, for example, the hypothesis that natural 

 selection is the dominant factor affords any explana- 

 tion of the fact that the numbers and proportions of 

 local species increase towards the south ; nor how it 

 can explain the fact that in New Zealand (c/. Ann. of 

 Bot., vol. 32, 1918, p. 339) a great many famiUes 

 show their maximum number of endemics in every 

 genus at the far north, all these families being Indo- 

 Malayan ; while a second group of families, character- 

 istic of the northern hemisphere, show their maximum 

 number at the south of New Zealand, and a third 

 group at the centre. The northern famiUes and 

 genera diminish as one goes southward in New 

 Zealand, and pass over, without paying any attention 

 to, the regions where the maxima of the central 

 and southern groups occur. These groups in the 

 same way show no unusual change when they reach 

 the region where the northern maximum occurs. Are 

 the environmental conditions so peculiar at these 

 points that those of the north should cause a multi- 

 plication of species only in Indo-Malayan families, 

 and those of the south only in families of the northern 

 hemisphere ? j. C. Willis. 



G. Udny Yule. 



Columnar Structure in Sandstone Walls of a Glass 

 Furnace. 



In the issue of Nature for December 29, 1921, 

 p. 567, I described the occurrence of columnar struc- 

 ture in optical glass and in fireclay. 



Through the courtesy of Mr. Currie, of the Scottish 

 Central Glass Works, Alloa, I had recently an oppor- 

 tunity of examining columnar structure that had 

 developed m the lowest sandstone course of the side 

 walls of a small tank glass furnace. The walls com- 

 prised two upper courses of fireclav blocks in which 

 no columnar structure developed,' and the bottom 

 course of rough-grained sandstone blocks obtained 

 from the Penshaw Quarries, Durham. Their cross- 

 section was about i sq. ft. Firebrick jack-arching 

 formed the floor of the tank, under which was 

 situated the regenerator. 



The sandstone course was laid in August 19 13 and 

 taken down in November 1921, during which opera- 

 tion the structure was observed. 



When emptying the tank the floor failed, and the 

 glass discharged itself through a space between the 

 floor and the regenerator roof. Thus while the walls 

 were rapidly chilled the floor was maintained at a 

 comparatively high temperature. 



The accompanying photograph (Fig. i) is of one 

 typical fragment taken from the inner surface ; other 

 portions showed curvature of the columns, which at 



the upper end were nearly normal to the corroded 

 A-shaped surface of the joint, and at the lower to 

 the bottom surface of the block. 



The similarity between these sandstone specimens 

 and those of optical glass previously illustrated is 

 worthy of remark. James Weir French. 



Anniesland, Glasgow, February 13. 



NO. 



2731, VOL. 109] 



The Action of Sunlight : A Case for Inquiry. 



Readers of Nature are no doubt aware that the 

 Medical Research Council has just appointed a 

 Committee on the action of light upon the human 

 body in health and disease, thus meeting the need 

 which I have been allowed to urge in these columns 

 under the above heading (Nature, December 8, 1921, 

 and January 5). 



I see no end to the inquiries in which we are now 

 at last to participate in England, the country the 

 smoke-darkened cities of which need them most. 

 Before me now is a series of papers which I owe to 

 Dr. A. F. Hess, of New York, who has demonstrated 

 that sunlight can cure or prevent rickets in human 

 infants and animals irrespective of the absence or 

 presence of the supposed an ti -rachitic vitamin. 

 Again, along this coast, from Cannes to San Remo, 

 I find French and Italian clinicians at work curing 

 what I have called the diseases of darkness by sun- 

 light ; also a voluminous literature, as yet entirely 

 unknown in England, which raises questions of high 

 racial, genetic, and eugenic importance, such as the 

 influence of sunlight, or the lack of it, upon the normal 

 development of the reproductive system and its 

 functions during adolescence. But clinicians else- 

 where had assured me — and I fear I may have 

 repeated their statements in these columns — that the 

 sun-cure cannot be practised on the Riviera ! 



Never henceforth, I predict, will the columns of 

 Nature cease to bear records of the new study of the 

 biology of light now to be begun. C. W. Saleeby. 



Hotel Royal Westminster, Menton, 

 February 19. 



