June i6, 1921] 



NATURE 



489 



of selective radiation-pressure alone is capable of 

 neutralising the force due to the gravitational attrac- 

 tion of the sun, but it looks very nnuch as if this 

 were so. Without being dogmatic on this point, we 

 can work out the consequences of this assumption. 

 In the case of stars having a much larger surface 

 temperature, say 14,000° K, B8A class, the value Ex 

 for H- and K-light would be much larger, so that the 

 radiation-pressure is still greater, and in some cases 

 preponderates over the greater value of gravitational 

 force on these stars. Thus Ca+-atoms would be 

 driven very far into the surrounding space. They will 

 be prevented from absolutely leaving the system, be- 

 cause with increase of distance the solid angle sub- 

 tended by the disc of the star at the atom woulo 

 diminish, and a condition of equilibrium would at last 

 be reached. 



The same phenomenon occurs to a smaller extent, 

 in the case of the sun, with Sr+ and Ba+, which have 

 their resonance-lines near the spectral region of maxi- 

 mum intensity, but owing to their greater atomic 

 weight the compensation is not so marked. Still Sr+ 

 is very prominent in the chromospheric spectrum, 

 rising to a height of 6000 km. 



The question may be asked : Why do we not obtain 

 the same phenomenon in the case of the other light 

 elements? These can be divided into two broad 

 groups : (i) non-metals like H, He, N, O, Ne, and 

 A. having a high ionisation-potential, of which the 

 resonance-lines lie in the extreme ultra-violet — e.g. for 

 H, at A=i2i6 A.U. ; for He, at A = 585 A.U. (Lyman 

 and Fricke, Fhil. Mag., May, 1920) — and can be de- 

 tected only by subordinate lines — for helium, by D3, 

 2p — md; for hydrogen, by the Balmer lines. Natur- 

 ally the effect of selective radiation-pressure is small 

 on these elements. (2) Elements like Na, K, Mg, 

 Al, Sc, Ti, Fe, which have an ionisation-potential 

 varying from 5 to 8 volts. Under the conditions 

 treated here these are mostly ionised, but the reson- 

 ance-lines of these ionised elements lie mostly out- 

 side the region available for observation, e.g. the 

 resonance-lines of Mg+ are A = 27955, 28027. The 

 resonance-lines of Na+ and K+ have not yet been dis- 

 covered, and probably lie in the extreme ultra-violet. 

 Sc+ and Ti+ are represented by prominent lines in the 

 chromospheric sp>ectrum, but it is not yet known 

 whether these are resonance-lines of these elements. 



The hypotheses thus appear to be promising, but 

 nothing final can be said before we can calculate the 

 absolute value of the selective radiation-pressure on 

 an atom. According to Eddington (Monthlv Notices, 

 R.A.S., 1920, vol. Ixxx., p. 723), the absolute value 

 of the radiation-pressure is too small to account for the 

 total neutralisation of gravitational force on the sun ; 

 but in that paper the consequences are worked on 

 the basis of the continuous theory of light. The fore- 

 going line of investigation at least brings out the 

 intimate connection between the stationarv character 

 of the H- and K-lines in the space round the stars 

 and the great prominence of these lines in the chromo- 

 spheric spectrum. It shows that the higher chromo- 

 spheric levels, as well as the space round B- and A-stars, 

 may probably contain, besides Ca+, also Na+, K+, 

 Sc+, Ti+, and Mg+, but owing to the fact that our 

 observations have to be limited between A = 3ooo A.U. 

 and 6000 A.U., and that none but the resonance-lines of 

 Ca+ He within this region, we can detect nothing but 

 Ca+. But if some day we can overcome the limita- 

 tion imposed bv atmospheric absorption, probably we 

 shall be able to detect Li+, Na+, Mg+, K+ in the 

 atmospheres surrounding B-stars which show 

 stationarv H- and K-lines. Megh Nad Saha. 



Berlin.' May 8. 



NO. 2694, VOL. 107] 



Biological Terminology. 



My gentle touch has started an avalanche indeed, 

 but I remain unmoved. Sir Archdall Reid asks 

 (Nature, June 2, p. 425) : " Is not all systematic 

 zoology and botany founded or^ this kind of classifica- 

 tion?" — a classification based on definite, concrete 

 facts of structure, in which there is "little or 

 nothing " based on causes, on antecedents and con- 

 sequents, or on hypothesis. The answer is in the 

 negative. May I illustrate briefly some kinds of 

 interpretation that a systematist has to employ? 



There lie before me some mushroom-shaped objects 

 from the Permian of Timor, clearly echinodermal, and 

 actually described as the swollen spines of a sea- 

 urchin. Such a spine is normally attached to the 

 shell of the urchin by a ball-and-socket joint. These 

 bodies, however, present at the end of the stalk three 

 articular facets, each with a straight fulcral ridge, 

 so placed that the fulcral ridges form an approxi- 

 mately equilateral triangle. Now, setting all resem- 

 blances aside, it is obvious that a single appendage 

 cannot be attached to an immobile base by three facets 

 so disposed, because the result of such an arrange- 

 ment is immobility. It follows, from equally clear 

 mechanical principles, that each facet must itself have 

 borne a single appendage. Consequently the mush- 

 room-like body is not an appendage, but a base which 

 once bore three appendages. In short, it must be 

 the cup and base of a crinoid. Having reached this 

 conclusion by the application of mechanical principles, 

 one attempts to apply some test, even if not a crucial 

 test in the strict sense. The stereom of a spine is 

 relatively light, and the meshwork in the axial region 

 is still more open ; the stereom of a fused crinoid 

 base is dense. Sections across the Timor fossil show 

 that its stereom is of the latter character. Not until 

 all the facts have thus been interpreted can we pro- 

 ceed to apply the methods of a postal address and 

 deliver our fossil at its proper street and number in 

 Crinoid-town. 



But there are cases in which the address is almost 

 illegible, or has been so often crossed through and re- 

 written that recourse must be had to skill higher than 

 that of a letter-carrier. I am at the moment trying 

 to identify some fossil Blastoids from North America. 

 Of recent years the rocks in which these genera are 

 found have been so minutely subdivided and the species 

 have been so finely discriminated that the ordinary 

 descriptions and keys (postal directories) cease to be 

 of much help. In this class, as in others, the same 

 forms appear to recur at intervals of time, and a 

 correct interpretation demands a close study of the 

 development in correlation with the chronology ; by 

 applying, as others have done, the theory of recapitula- 

 tion we may unravel the tangle. It is not only fossils 

 that furnish such problems to the philosophic inter- 

 preter; Dr. Annandale was showing me yesterday 

 some Gastropods from Asiatic lakes that have to be 

 dealt with in just the same wav. 



If we turn to the broader divisions of systematic 

 zoology we derive still less aid from those simple 

 rule-of-thumb methods which represent to Sir Arch- 

 dall Reid the principles of taxonomy. At everv step 

 the modern systematist is considering origins ; for 

 him the truth or falsitv of such principles as "the 

 irreversibility of evolution " is of vital importance ; 

 his very diagnoses embody speculations. But the 

 svstemsitist recognises the metaphysical nature of his 

 classifications, and he is perpetually seeking some 

 crucial instance that shall give them a more secure 

 basis of fact. He prophesies, for examole, the exist- 

 ence of some connecting type at a certain period, and 

 then he goes and finds it. 



So much for the systematist ! As for the biologist 



