NA TURE 



301 



THURSDAY, NOVEMBER 4, 1920. 



Editorial and Publishing Offites : 



MACMILLAN &• CO.. LTD.. 



ST. MARTIN'S STREET. LONDON, W.C.2. 



Advertiiemems And business letters should be 

 addressed to the Publishers. 



Editorial communications to the Editor. 



Telegraphic Address : PHUSIS. LONDON. 

 Telephone Number: CERRARD 8830. 



Symbolic Language of Science. 



WE have received from Prof. McAdie, of Blue 

 Hill Observatory, an off-print of a note 

 on the subject of uniformity in aerographic nota- 

 tion, which occupies nine pages (169-177) of the 

 Blue Hill Meteorological Observations, forming 

 vol. Ixxxiii., part 4, of the Annals of the Harvard 

 Observatory. It invites agreement with regard 

 to the symbolic representation of the quantities 

 and operations required for the discussion of 

 problems in the dynamics and physics of the 

 atmosphere in pursuance of a suggestion of. 

 Dr. Otto Klotz in Science, vol. xlvi., p. 360, 

 1917. 



There can scarcely be any difference of opinion 

 as to the desirability of arriving at an agreed 

 practice in the use of symbols by different workers 

 in the same subject. So far as the atmosphere 

 is concerned. Prof. Bigelow held a very compre- 

 hensive review of the rank and file of the analysis 

 of atmospheric operations in his discussion of 

 cloud observations which forms the second 

 volume of the Report of the Chief of the Weather 

 Bureau for 1898-99; the first volume of the 

 work on " Dynamic Meteorology and Hydro- 

 NO. 2662, VOL, 106] 



graphy," by V. Bjerknes and others (Carnegie 

 Institution of Washington, 1910), sets out the re- 

 presentation of the physical quantities to be 

 employed in a systematic manner, and a book by 

 Mr. L. F. Richardson, now in the printer's 

 hands, ranges over the same field. Ideas of uni- 

 formity are naturally developed to some extent in 

 the various parts of the Computer's Handbook 

 of the Meteorological Office, which endeavours to 

 provide the worker in atmospheric dynamics and 

 physics with the necessary material and forms for 

 numerical computation arranged upon a sys- 

 tematic basis, and aspires to attach to each 

 quantity its appropriate symbol. 



But the subject bristles with difficulties of many 

 kinds. We have available in practice, let us say, 

 five alphabets ("lower case" roman and italic, 

 "upper case" roman and italic, "lower case" 

 Greek), a few letters taken from other alphabets, 

 and a few additional symbols for units or opera- 

 tions. With these we have to represent the items 

 of a considerable number of quite separate cate- 

 gories ; for example, we want to represent: (i) 

 defined qnantities for specific units, as m for 

 metre, g for gram; (2) quantities undefined as 

 to unit, as <f> for latitude and X for longitude, 

 which may be in radians or degrees, and u, v, w 

 for velocity irrespective of unit; and (3) symbols 

 of operation, as of multiplication, division, sum- 

 mation, and differentiation of various kinds. 



In the absence of a sufficiency of letters or 

 alphabets, we are accustomed to make out with 

 suffixes and indices, but here we are liable to get 

 into vexatious difficulties with the printer, and 

 more particularly with the typist, who must be 

 reckoned with in these days, and is apt to display 

 a misguided ingenuity in substituting for a care- 

 fully selected symbol the nearest thing to it that 

 can be got out of the keys ; and, in any case, if 

 the symbolism could be so arranged as to allow 

 the print to "run on," it would be useful for both 

 reader and publisher. 



No adequate account can be given of the details 

 of Prof. McAdie's scheme without reproducing it 

 in extenso. With three alphabets and some 

 suffixes and indices (apparently without making 

 any use of the discrimination between roman and 

 italic type), he represents ninety-four quantities, 

 including such modern conceptions as electron, 

 Planck's element of action, Taylor's eddy conduc- 

 tivity and viscosity, and Richardson's turbulivity. 

 He uses them in five pages of fundamental re- 

 lations. But the allocation is not altogether sue- 



