September 28, 191 1] 



NATURE 



411 



historical times. This number, it will be observed, 

 not correspond with the one just given. Although 

 we observe certain lacuna; in this catalogue, it is a 

 very useful compilation for those interested in 

 vulcanology. 



Although geologists may object to Dr. Schneider's 

 new terminology, they must not overlook the fact 

 that they themselves have had many christening 

 parties. Some thirty or forty years ago a distin- 

 guished president of the Geological Society, when he 

 first heard of Belenites — a word not unlike Beloniten — 

 interrupted the speaker by the remark, "Tut! tut! 

 Belemnites in volcanic rocks indeed!" If we say 

 nothing about the new language, then Dr. Schneider 

 is to be congratulated on his work. John Milne. 



BIOMETRICAL METHODS FOR 

 PSYCHOLOGISTS. 

 The Essentials of Mental Measurement. By Dr. Wm. 

 Brown. Pp. vii+154. (Cambridge: University 

 Press, 191 1.) Price 35. 6d. net. 



THIS volume is based on a research "devised for 

 the purpose of determining to what extent cor- 

 relation exists between certain very simple mental 

 abilities, in cases where the individuals experimented 

 on are, as near as may be, identically situated with 

 respect to previous practice, general training, and 

 environment; and how closely, if at all, these ele- 

 mentary abilities are related to general intellectual 

 ability as measured by teachers' judgments, school 

 marks, &c." 



The groups of individuals tested were homogeneous 

 as far as possible, and were selected from a London 

 elementary school, a London higher grade school, 

 from among women students of a training college, 

 and men and women university students. 'I he 

 tests employed were fourteen in number. In two of 

 them the subject was required to cross out as quickly 

 and accurately as possible on a printed page certain 

 letters (e and r in one case, and a, n, o, s in the other), 

 in .1 third all the letters were to be crossed out. Speed 

 and accuracy in the addition of single digits were 

 measured in a fourth ; while in the fifth, sixth, and 

 seventh the ability to estimate the relative lengths of 

 printed lines was tested in different ways, including 

 conditions under which illusions are known to be pro- 

 duced. The power of memorising poetry and strings 

 of nonsense syllables were examined in the eighth and 

 ninth, while the tenth test was the combinations 

 methode of Ebbinghaus. In this "the subject is 

 shown a passage of continuous prose with one-third 

 to one-quarter of the words replaced by blanks, and 

 is asked to supply the missing word or words of 

 similar significance." The eleventh and twelfth were 

 marks obtained in the ordinary school curriculum for 

 drawing, and the total school marks ; in the thirteenth 

 general intelligence was graded by the independent 

 estimates of two different teachers, which were found 

 to be very closely correlated together. In the four- 

 teenth association time was measured by counting the 

 number of words associated with a given word which 

 the subject could write down in a given time. 



The majority, but not all the tests, were applied 

 NO. 2187, VOL. 87] 



throughout each of the six groups, and the statistical 

 treatment of the records is very complete and satis- 

 factory. The results for each group were, of course, 

 kept distinct, and correlation coefficients were calcu- 

 lated between the result of each test and the result of 

 each of the others, while to estimate the trustworthi- 

 ness of each the correlation between the results of two 

 separate applications of the test on each subject was 

 measured. In addition to this, certain groups of tests 

 were taken and the partial correlation coefficients cal- 

 culated, i.e. the correlation between any two members 

 of a group for constant values of the remaining mem- 

 bers. As the probable errors are given for all the 

 coefficients calculated, comparisons between them can 

 be satisfactorily made. 



Limitations of space forbid us from giving an 

 account of all the conclusions drawn by the author, 

 but two at least can be stated very simply and are 

 worthy of note; firstly, that the combinations methode 

 of Ebbinghaus is a good measure of mental ability. It 

 ' correlates with "general intelligence" almost as 

 clearly as "scholastic intelligence" (school marks) 

 "does"; and secondly, mechanical memory, i.e. the 

 power of memorising a number of nonsense syllables, 

 is also fairly closely correlated with general intelli- 

 gence. This research was published in the British 

 Journal of Psychology for October, 19 10. Together 

 with the chapters which appear in this volume on the 

 mathematical theory of correlation and on the history 

 of the use of the theory in psychology, it formed a 

 thesis approved for the degree of doctor of science in 

 the University of London. The thesis, with chapters 

 added on mental measurement, psycho-physical 

 methods, and on the significance of correlation in 

 psychology, constitutes the bulk of the present volume. 

 Its object appears to be not so much to instruct the 

 beginner in psychology on the best way of measuring 

 psychical characters, but to introduce the professed 

 psychologist to the biometrical methods of Prof. Karl 

 Pearson and his school. The author shows a consider- 

 able grasp of the subject, and does not perhaps realise 

 the difficulties of less mathematically minded people 

 than himself. Thus the chapter on the mathematical 

 theory of correlation, if intended for an introduction 

 to the subject, could well have been expanded, ren- 

 dered more elementary, and more amply provided with 

 examples. It is by examples rather than by precept 

 that people who are confused by algebraical symbols 

 are able to learn to use the methods. Those who can 

 readily follow the notation will find the book profit- 

 able to read and useful for reference, particularly as a 

 good bibliography is appended. E. H. J. S. 



ELECTRIC CRANES. 

 Electric Crane Construction. By Claude W. Hill. 



Pp. xx + 313. (London: C. Griffin and Co., Ltd., 



1911.) Price 25s. net. 



\ LARGE measure of flexibility in the supply of 

 -V power is an essential feature in all hoisting 

 machinery. Not only must the point from which the 

 load is suspended be capable of movement in different 

 directions, but the whole machine must be able to 

 travel, and these requirements render electricity a par- 



