NATURE 



THURSDAY, MAY 22, 1913. 



THE ROYAL SOCIETY'S SUBJECT INDEX. 

 Royal Society of London. Catalogue of Scientific 

 Papers, 1800-1900. Subject-Index, Yol. iii., 

 Physics. Part i., Generalities, Heat, Light, 

 Sound. Pp. c + 550 + vii. (Cambridge Uni- 

 versity Press, 1912.) Price 185. net. 



THE most obviously essential qualities of a 

 work of reference such as this are accuracy, 

 comprehensiveness, and a lucid arrangement of 

 the contents, so that anyone using the book may 

 readily find the information he is in search of. 

 The test of accuracy must lie in the result of long 

 usage, but, in the present case, so far as an 

 a priori guarantee can go, we have it 

 in the auspices under which the book is 

 produced. As to comprehensiveness, while abso- 

 lute completeness is no doubt unattainable, most 

 of those who have occasion to consult the work 

 before us will be satisfied with the statement that 

 it "contains 33,344 entries referring to the papers 

 contained in 1261 serial publications." Even 

 these numbers convey little idea of the compre- 

 hensive scale of the whole book of which the 

 volume we are considering forms a part, unless 

 it is borne in mind that it deals with onlv 

 those parts of the science of physics which are 

 included under the headings generalities, heat, 

 light, sound, leaving the great subjects electricity 

 and magnetism for another volume, and unless we 

 remember, further, that the whole of physics con- 

 stitutes only one of seventeen sciences included 

 in the schedule of the International Catalogue. 

 These sciences are mathematics, mechanics, 

 physics, chemistry, astronomy, meteorology, 

 mineralogy , geology, geography, palaeontology, 

 biologv, botany, zoology, anatomy, anthropology, 

 physiology, and bacteriology. 



The volume relating to pure mathematics was 

 published in 1908 and contains 38,748 entries 

 referring to 700 serials, and the volume on 

 mechanics, published in 1909, contains 21,293 

 entries referring to the papers contained in 959 

 serials. This makes a total of 93,387 entries con- 

 tained in the three already-published volumes of 

 the catalogue, or, if we assume that the second 

 half of " Physics " will yield as many as the first 

 half, we get an estimated total of more than 

 126,000 entries for the first three sciences in the 

 above list. Of the remainder, some will no doubt 

 provide a smaller number than the average of 

 those already dealt with, but others will probably 

 furnish quite as many. This may suffice to give 

 some idea of the comprehensive character and 

 NO. 2273, VOL. 91 ] 



immense scale of the work which the Royal 

 Society's committee has undertaken in the com- 

 pilation of this catalogue. 



The arrangement of the matter has obviously 

 required very careful consideration. A purely 

 alphabetical arrangement of such an immense 

 number of entries relating to such a great variety 

 of subjects would clearly have resulted in a series 

 of bewildering lists wherein the search for a 

 particular item would have been like that for a 

 needle in hay. The arrangement actually adopted 

 is founded on an elaborate attempt at a rational 

 classification of the subject-matter of the sciences 

 dealt with. This is carried out by a series of 

 successive divisions and subdivisions, the nature 

 of which can be best shown by an example. Thus 

 the general subject of Heat is first distributed 

 among the following main divisions : General ; 

 Sources of Heat and Cold ; Thermometry ; Rela- 

 tions involving Expansion and Stress ; Calon- 

 metry and Specific Heat; Phenomena of Change 

 of State; Thermal Conduction and Convection; 

 Thermo-dynamics. Each of these main divisions 

 is subdivided into numbered headings; thus, 

 to take a comparatively compact example, 

 Calorimetry and Specific Heat comprises 

 the following headings : 1600, General, Units 

 of Heat; t6to, Calorimetric Methods; 1620, 

 Specific Heats of Solids and Liquids ; 1640, Speci- 

 fic Heats of Gases and Yapours; 1660, Chemical 

 Constitution and Specific Heat; 1670, Heats of 

 Eusion; 1680, Heats of Vaporisation ; 1690, Heats 

 of Dissolution ; 1695, Heats of Transformation. 



The numbers greatly facilitate cross-reference. 

 It will be seen that they do not run continuously 

 and that the intervals between them are n.oC 

 uniform. These intervals make it possible to 

 expand the index in future by inserting additional 

 entries without disturbing those to which numbers 

 have already been assigned. Eor some father 

 occult reason the reference numbers all have four 

 figures, the first half-dozen being 0000, ooi/o, 0020, 

 0030, 0032, 0040. These numbered headings are 

 in most cases again subdivided, sometimes to a 

 considerable extent, before we come to the refer- 

 ences to individual papers. 



The extreme terms of this serines of divisions 

 and subdivisions, namely, a brancjh of science and 

 a particular paper relating to sorrie matter falling 

 under this branch, are determined by the nature 

 of the case; but there is room for almost any 

 amount of difference of opinion as to how many 

 intermediate terms should be interposed, and 

 where they should be placed. The ultimate 

 criterion in this matter should be, in our opinion, 

 the degree of ease and convenience with which a 



