July 20, 1916] 



NATURE 



419 



of probabilities, the weighting- of observations, 

 and the treatment of the figures obtained, but 

 illustrations are deferred until after the chapter on 

 the precision of observations. Some interesting 

 subjects for discussion are appended. The last 

 relates to gambling, and the views of Dr. Burn- 

 ham, of Chicago, are quoted, who believed .that 

 if the laws of chance were taught to children in 

 the schools, they would steer clear of the slot 

 machine in early years, and later would shun the 

 bookmaker and every other gambling magnate. 

 Now, would they? Might not they, even though 

 they had been taught that the value of the chance 

 was .only halt what they were paying, come to 

 that other conclusion — natural if they have im- 

 perfectly understood what they were taught — that 

 the laws of chance are "all theory like the stars," 

 and that with luck they might easily win a big 

 prize ? 



The third chapter is on "the adjustment of 

 observations,' and here we find more pains taken 

 to explain how observations in general and ob- 

 servations that are not exactly consistent in par- 

 ticular should be dealt with to obtain the best or 

 most likely results. 



In the chapter on "the precision of observ^a- 

 tions " the probability curve is treated graphically, 

 and mean square error, average deviation, and 

 probable error are explained. The next chapter, 

 on the propagation of errors, perhaps most nearly 

 touches the experimental work of the student, for 

 here the relation of error of observation to error 

 of result is discussed. After this, plotting and 

 negligibility are the subjects of two chapters, in 

 the latter of which the slide rule is taken as an 

 example. The concluding chapter is on empirical 

 formulae and constants. 



It will be seen from the tabular statement of 

 the subjects considered that they are of the first 

 importance to the experimentalist. At the same 

 time, unless the student is made to appreciate 

 well both the niceties of the experimental art and 

 the matters dealt with in this book, the latter 

 may, if imperfectly understood, be a source of 

 danger. The student may not appreciate the 

 futility of overloading a multitude of bad observa- 

 tions, subject of necessity to consistent errors, with 

 sheets of least square calculations. If he has more 

 aptitude for figures than for experiment, he may 

 even delude himself into believing that his cal- 

 culated probable errors really are probable errors. 

 In such cases it is much more important to spend 

 the time required for these calculations in im- 

 proving his apparatus or varying his method so as, 

 so far as possible, to avoid consistent errors. Two 

 or three experiments really well conducted are 

 worth far more than a multitude performed in a 

 slovenly wav, and no scientific juee-ling will give 

 the multitude more value. The writer feels that this 

 aspect of the general question is not sufficiently 

 insisted on, and the book, in spite of its many 

 excellent features, would be more valuable to the 

 student if the author had condescended to give 

 more attention to the actual operations of the 

 laboratory and their relation to the consequent 

 calculations. C. V. Boys. 



NO. 2438, VOL. 97] 



■■srcrrrED fever. 



Cerebrospinal Fever. By Dr. Michael Foster 

 and Dr. J. F. Gaskell.' Pp. x + 222. (Cam- 

 brijlge: At the University Press, igi6.) Price 

 125. 6d. net. 



'T'HIS excellent and complete monograph of the 

 ^ much-dreaded disease, cerebro-spinal fever — 

 dreaded because of its high mortality and inca- 

 I>acitating sequelae — should prove of great interest 

 not only to the members of the medical profession 

 but to men of science generally. 



The book is dedicated by the authors to the 

 memory of their respective fathers, and on account 

 of its careful, lucid, scientific, yet withal practical, 

 exposition of the subject it is a worthy tribute 

 to those two great founders of the modern school 

 of English physiology. 



The authors claim that this monograph has for 

 its aim an attempt to bring together and correlate 

 the clinical and pathological facts which they were 

 enabled to accumulate during the epidemic of 1915 

 in the Eastern Command, and the views set forth 

 are the outcome. of clinical and pathological obser- 

 vations made in the wards, the laboratory, and 

 tlie post-mortem room of the ist Eastern General 

 Hospital. 



There are eleven chapters and two appendices, 

 and the excellent plan of giving a summary in 

 italics of the principal facts dealt with in each 

 chapter is helpful to the reader. There are eleven 

 excellent plates, eight of which are coloured. 

 The work commences with an interesting histori- 

 cal account of the disease — largely a summary 

 from the exhaustive treatise by Hirsch on Geo- 

 graphical and Historical Pathology. The first 

 authentic account of an epidemic is that which 

 occurred in Geneva in 1805. From the date of 

 this, its first appearance, the disease was epidemic 

 at various places both in Europe and .America. 

 Read in the light of modern knowlege of carriers 

 in the propagation of disease, we can understand 

 how this disease suddenly appeared and travelled 

 according to no appreciable law. 



Prior to 191 5 cerebro-spinal fever in an epi- 

 demic form had been confined in Great Britain to 

 the industrial centres of Scotland and Ireland. The 

 authors point out that although the naso-pharynx 

 is the location in which the specific organism is to 

 be found, yet, according to their experience, it may 

 be present without causing any marked inflamma- 

 tory condition of the mucous membrane. Conse- 

 quently, carriers may appear to be healthy per- 

 sons; and it is not surprising, therefore, that 

 when, in 1915, large numbers of soldiers were 

 crowded into huts and billets with deficient venti- 

 lation and other favouring conditions, outbreaks 

 of the disease should have occurred not only 

 among the soldiers but also among civilians. 

 Serious epidemics occurred at Salisbury Plain, 

 Aldershot, in the London area, and in the eastern 

 counties of England. A good account of the 

 symptomatology, diagnosis, and treatment of the 

 disease is given. Four excellent coloured plates 

 illustrate the four distinct varieties of rash, and 



