302 



NATURE 



[NOVEMFER 4, 1920 



cessful; the letters are all gone, and some very 

 ordinary quantities, such as area and vapour- 

 pressure, have no symbol. N means Avogadro's 

 constant, S entropy, E energy, and W external 

 work ; but somehow or other, definition or no de- 

 finition, N— S and W — E are written down as re- 

 presenting wind components. If it be agreed 

 that kb means kilobar and s means space, there 

 is a certain depravity in using kbs, as Prof. 

 McAdie does, to mean simply a plural of kb. 



Furthermore, it is not really possible with our 

 limited number of letters to allocate so many of 

 them for purely arithmetical purposes. Very many 

 of them are wanted for algebra; in fact, many 

 more than are actually available before any allo- 

 cation takes place. Prof. McAdie allows only 

 X, y, s, and t for co-ordinates, but we must be 

 permitted to change our co-ordinates. In the 

 past, polar co-ordinates, co-ordinates referred to 

 moving axes, and direction cosines have claimed 

 letters which Prof. McAdie proposes to hand over 

 for fixation. 



The fact is that the allocation of a letter in 

 common use to one specific and conventional 

 meaning is a very awkward policy. It is not 

 unfair to argue that if we want a symbol to be 

 allocated definitely to a new quantity, and taboo 

 for everything else, we ought to invent a new 

 figure and get it canonised by a sufficient 

 authority to induce printers to stock it. It sounds 

 impracticable, but is at least worth consideration, 

 for meteorologists have already done it in the case 

 of symbols for atmospheric phenomena. Other- 

 wise workers in sciences which are adjacent and 

 not altogether independent are likely to select the 

 same letters for different quantities, and there 

 must be confusion. Perhaps the language of 

 science ought to expand its notation as ordinary 

 language has done by proceeding from single 

 letters to syllables, and that would certainly be 

 an easy and effective way of dealing with the 

 question if we could do away with the convention 

 that multiplication needs no symbol of operation, 

 and require that every operation should be repre- 

 sented by a suitable sign. 



Indeed, all ideas of allocating symbols for 

 special quantities lead up to the suggestion that 

 the study of the effectiveness of the language of 

 science ought no longer to be left to the casual 

 play of forces of individual workers. An Academy 

 of Science might with advantage have a literary 

 side, and there might even be lectures on the art 

 of expression, symbolically or otherwise, of scien- 

 Np. 2662, VOL. 106] 



tific truths, with examples, some good and some 

 bad, taken from scientific literature. Scientific 

 institutions tend to separate themselves from the 

 study of the classical languages and become inde- 

 pendent centres of learning, and as they do so 

 they ought to make adequate provision for the 

 study of the history and literature of science in 

 order to make sure that the literary form of the 

 results which the institutions will present to the 

 world will not be entirely neglected, as apparently 

 it is now in some contributions to science. Scientific 

 achievement is not really complete when only the 

 author is satisfied with the results. It might be 

 useful to make out how much scientific literature 

 is neglected because its form is crabbed, repul- 

 sive, or even unintelligible. 



These are quite natural expansions of the idea 

 of economising the effort of the writer and reader 

 by an agreement as to the use of symbols. That 

 is an obvious and essential step, but we ought 

 first to come to an agreement upon some general 

 principles whether, to begin with, there is, or shall 

 be, any difference in meaning between a symbol in 

 roman and the same letter in italic. The Com- 

 puter's Handbook draws a distinction. It re- 

 serves roman for units or symbols of operation. 

 It uses italic to indicate varying quantities, and 

 sometimes capitals for defined quantities. Thus g 

 is gram, but g is gravitational acceleration ; is 

 temperature, but T is temperature on the absolute 

 scale. Another general question is whether opera- 

 tions should be consistently indicated by letters 

 or by signs, or either mode of indication be 

 allowed as the writer finds convenient. 



The difficulty in the way of getting agreement 

 on these questions is not so much unwillingness 

 on the part of workers as lack of authoritative 

 reference to recognised classics. After carrying 

 on an elaborate series of computations with the 

 improvised symbolism that suggests itself as one 

 goes along, it is too laborious to go over it again 

 to bring it into agreement with someone else's 

 notions ; and as one is naturally led on gradually 

 in the course of a research to take in more quanti- 

 ties, one can scarcely begin it by selecting a system 

 of notation. It grows of itself as the work pro- 

 gresses. But if the Royal Society were to alter- 

 nate the list of publications as an advertisement in 

 the lining of its Transactions with a list of recog- 

 nised symbols, or if in some other way some con- 

 vention were made quite easy of reference, we 

 should have little difficulty in adhering to it. 



Napier Shaw. 



