242 



NATURE 



[January i6, 1902 



from them a psychological theory of the successive evolu- 

 tion of grammatical form and syntactical construction. 

 This task is performed in the two chapters which deal 

 with "word-forms" and "sentence-construction." 

 These are followed by a chapter on the psychical causes 

 of change of signification, and the whole work concludes 

 with a final chapter devoted to a brief rc'siim^ and 

 criticism of the various theories of the " origin of 

 language." It is significant of the influence exerted 

 upon modern psychological thought by the concept of 

 evolution that this last chapter is by far the shortest in 

 the book and that its result is in the main the purely 

 negative one that speech, like the other human faculties, 

 has no definite beginning or origin, but is connected by a 

 continuous development with a pre-articu!ate and pre- 

 human past, while the earliest stages of language known 

 to us themselves presuppose a long development within 

 human speech itself. Of the elaborate investigations 

 which fill Prof Wundt's chapters vi.-viii. it is impossible, 

 for reasons of space, to say much e.xcept that they are of the 

 highest psychological interest and importance. Specially 

 important are the distinction between a preeminently 

 nominal and objective type of language, like those, e.g., 

 of the American or of the Ural-.A.ltaic family, and a pre- 

 eminently verbal and subjective type, like that to which 

 we are accustomed in the familiar Indo-Germanic group, 

 and the very similar distinction, in the realm of syntax, 

 between the attributive and predicative types of pro- 

 position as corresponding to the "nominal" or objec- 

 tive and the "verbal" type of thought respectively. 

 These and similar differences, obviously pointing to 

 marked divergence of psychological endowment, are 

 minutely and learnedly discussed by Prof Wundt in a 

 way which only makes one regret that his weakness for 

 diffuse expression makes it so difficult to get a clear and 

 systematic grasp of his argument as a whole. 



A. E. T. 



ELECTRICAL ENGINEERING TESTING. 

 Electrical Engineering Testing. By G. D. Aspinall 

 Parr. Pp. viii-l-474; 218 diagrams, 31 tables. 

 (London: Chapman and Hall, Ltd., 1902.) Price 

 gj. net. 



DURING the last few years the rapidly spreading 

 use of electrical methods of dealing with engineering 

 problems has been naturally accompanied by the publica- 

 tion of a great mass of isolated work on the testing of 

 electrical materials and machinery. Successful design 

 requires not merely a knowledge of principles, but an 

 intimate acquaintance with the properties of materials 

 to be used ; that these may be used economically and 

 without needlessly high " factors of safety." And an 

 electrical engineer is required to understand how to test 

 for magnetic and electrical excellence the material he 

 may have to use, as well as how to test his machinery 

 for efficiency and staying power and his instruments for 

 accuracy. It is probable that electrical engineering 

 testing presents not only a possibility of, but a necessity 

 for, greater refinement of measurement than occurs in 

 other branches of engineering testing ; the success of 

 much of our electrical machinery and apparatus being 

 more immediately dependent on exact design. 

 NO. 1 68 I, VOL. 65] 



When a subject is developing so rapidly as that which 

 Mr. Parr has chosen, it is not easy to maintain an exact 

 knowledge even of what should be regarded as the 

 simpler matters, and it is still less easy to keep this 

 knowledge in due proportion. For that which was but 

 lately abstruse or useless may rapidly become clear and 

 even elementary information. 



The volume before us will prove useful to many, con- 

 taining, as it does, much information relating to elec- 

 trical testing which, in spite of a rather unsatisfactory 

 arrangement, is in a fairly accessible form. Its main 

 portion consists of accounts of about 130 different tests, 

 each arranged as follows : — A descriptive introduction ; 

 a list of apparatus wanted ; what observations to take 

 and how to take them ; and concluding with " infer- 

 ences " to be drawn from the results of the test. The 

 book is styled "a practical work," meaning, we presume, 

 not so much a text-book in which difficulties are ex- 

 plained as a hand-book containing useful information 

 about tests. It is primarily written for students, and 

 we do not doubt it will prove very useful alike to student 

 and demonstrator. 



The "descriptive introductions" are probably as satis- 

 factory as the nature of the book will allow. The list of 

 apparatus, while usually complete, will strike the reader 

 as relating in particular to the apparatus of Mr. Parr's 

 laboratory in the Yorkshire College. The next section 

 of each test on taking and tabulating observations is 

 very well given. It will save the student much thought 

 and keep him from much bungling, and will raise the 

 quality of his work — all but the first no doubt desirable 

 when the course must be hurried over. The " infer- 

 ences" are added at the end "to make the experimenter 

 think and reason for himself" If the test had not been 

 so well arranged for him beforehand, or the instructions 

 so fully given, perhaps he would have had to think for 

 himself before the test could have been carried out. 

 This is a matter of opinion ; but we think that the book 

 will prove more useful in evening classes than with day 

 students. 



An appendix gives in a few pages the derivation of 

 some formuhB employed, and the book concludes with 

 some 100 interesting pages descriptive of apparatus 

 used. The author shows much of his own apparatus 

 and methods, which are interesting even when they do 

 not strike one as the best. 



The book is nicely printed, but bears signs of haste in 

 preparation. Thus most readers, when they have 

 grasped the peculiar use of the word " inference," will 

 be surprised, if not amused, at the following : " Infer- 

 ence : Does the accuracy of the above test depend upon 

 anything in particular?" p. 17 ; and on p. 22, under a 

 similar heading, "Can anything in particular be deduced 

 from the above results?" The author spells converter 

 with an o in the final syllable, and writes anti-inductive 

 where non-inductive is meant ; and slips such as series 

 for serious, p. 154, sale for scale, p. 37S, Jolly for 

 Joly, p. 390, iVc, which are not infrequent, should have 

 disappeared in the proof. 



The figures are on the whole good, and the tables at 

 the end welcome. 



The author's style is often by no means lucid, and even 

 when the meaning is clear it does not make agreeable 



