January 25, 1900J 



NATURE 



295 



Brain in Relation to Mind. By J. Sanderson Chris- 

 tison, M.D. Pp. 142. (Chicago: Dr. Christison, 1899.) 

 " Comparatively few physicians and fewer laymen,'' 

 says Dr. Christison, "have a satisfactory view of the 

 relationship of brain to mind," and we are afraid that 

 Dr. Christison is not among the minority. When the 

 writer of a book that purports to be a serious work on 

 psychology commits himself to such statements as that 

 "the use of means to a given end implies the pre-exist- 

 ence of a specific potentiality having a plan in the 

 abstract, for only the pre-existing can be the cause of a 

 necessity"; and that "the evolution idea ... is abso- 

 lutely incompatible with law and order," it would be out 

 of place to deal with him m a serious review. 



LETTERS TO THE EDITOR. 



The Elitor does not hoLl himself responsible for opinions ex- 

 pressed by his correspondents. Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part of NATURE. 

 No notice is taken of anonymous communications .'\ 



Is New Zealand a Zoological Region .' 

 As some weeks must elapse before Mr. Farquhar is able to 

 reply to the questions asked (p. 273) by my good friend Mr. 

 Wallace, I would urge your readers to bear in mind that no 

 particular sanctity should be attached to the word " Region." 

 Such a sanctity was once ascribed to the word " Species," and 

 to no one living more than Mr. Wallace are we indebted for the 

 removal of the misconception therein involved. Properly used 

 both are very good words, and it would not be easy to get on 

 without them ; but to each may be ascribed a virtue which it 

 does not rightly possess. In the case of New Zealand I woukl 

 meanwhile refer to the expression of Prof. Huxley's views in 

 the Proceedings of the Zo ©logical Society for 1868, which so far 

 as I know have never been refuted. Alfred Newton. 



Cambridge, January 20. 



Compensation in Weather. 



As our climate does not, in a lonj;; series of years, undergo 

 permanent change, we are accustomed, in a time of extreme 

 cold, wetness, or other quality of weather, to expect, in a vague 

 way, that we shall have, ere long, a variation in the opposite 

 direction, equalising matters. Is it possible to attain greater 

 definiteness in such expectation, and, utilising the principle of 

 compensation, to forecast, in a measure, the character of an 

 approaching season ? 



I propose to inquire what sort of relation subsists between the 

 cold of a given winter and that of the thirty winters preceding. 



We may conveniently measure the cold of winter seasons by 

 the total number of frost days from September to May. (For 

 brevity, we shall here designate each winter season by the year 

 in which it ends : 1842, e.g. meaning 1841-42.) 



Let us, then, begin by adding the numbers of frost days in the 

 thirty winters 1842 to 1871, 1843 to 1872, and so on. We thus 

 get a series of twenty-nine numbers, varying from 1593 to 1717, 

 with an average of 1653. 



The following table may now be considered : — 



Groups of 30 winters 



containing, of 



frost days, 



r) Under 1610 



2) 1610-1629 



(3) 1630-1649 



(4) 1650-1669 



(5) 1670-1689 



(6) 1690 or more 



Thus, in general, the larger the number of frost days in the 

 30winter groups, the smaller, on an average, is the number in 

 the winter following. A reversal occurs at No. 5. (But there 

 are only two cases.) 



This is, perhaps, very much what we should expect on the 

 principle of compensation. 



Analysing further, we must bear in mind that the data 

 considered are not very numerous, and beware of building too 

 much on a slender foundation. The remarks here offered are 

 rather by way of suggesting a method which might perhaps be 



found practically helpful, if fuller data were found to point in 

 the same direction. We may, then, note these thiee facts : — 



(i) Thesix mildest winters (since 1 871) were each preceded 

 by a 30-year group having more than, the average of frost days. 



(2) Thesix coldest winters were each preceded by a 30 year 

 group having less than the average of frost days. 



(3) Of fifteen 30-year groups with excessive cold (i.e. over 

 the average), as many aj tiveive were followed by mild winters, 

 and only three {i.e. one-fifth) by severe winters. 



Applying the same method to frost days in the first half of 

 each year, it might be shown that out of twelve 30-year groups 

 with excessive cold, ten were followed by mild first halves, and 

 only two by severe first halves. 



Reverting to the former case (total winter seasons), the 

 diagram represents, in curve A, the total frosl days in 30- 

 winter groups ending 1871, 1872, &c. Below is an inverted 

 curve, the first point of which indicates the number of frost days 

 in winter 1872, the second in winter 1873, ^"d so on. (The 

 time scale below, it will be seen, is shifted one year to the left.) 



i^ '// > '7 'So '3 'C *cj W '6' '8 fqot 



^^ i I I I J I I I v<rv' 



NO. 1578, VOL. 61] 



V6 - 



'72. 



w^y^ 



<^v.Us'a. 



a.v. irr- 



f/^i: 



A. Curve of frost-days in 30 winters ending 1871, 1872, &c. 



B. Curve of frost-days in winters 1872, 1873, &c. (inverted). 



A certain correspondence may be traced between these curves. 

 And taking the dots above the average line in A, it will be 

 seen, how, except in three cases, marked by a small circle (in 

 B), each corresponds with a mild winter in B,a fact already noted. 



There is a curious lag in part of the second curve. We find 

 two conspicuous wave-crests in A, viz. in 1881 and 1893, ^"d 

 the highest points {i e. lowest values) in B are in 1884 and 1896. 

 Further, the lowest point in A is in 1885 ; the lowest in B in 

 1888. An interval of three years in each case. It might be 

 worth while to observe whether this is repeated to any extent 

 in the future course of these curves. 



Now the last thirty winter seasons (ending 1899) have a tota. 

 of 1657 frost days, and if we allowed ourselves to speculate as 

 to the quantity of frost in the current winter season (1900) we 

 might at least perhaps fairly anticipate, or might have antici- 

 pated, that this would not turn out an extremely severe winter, 

 and that the chances are rather against its being extremely mild 

 either. But the point I would emphasise is that, with copious 

 material, useful hints might possibly be obtained as to the limits 

 within which the character of a coming season (or other portion 

 of tinje), as to heat or other quality would probably occur. To 

 this end, I may remark, some might find a diagram of dots 

 helpful, in which, taking the case of winters, the abscissae and 

 ordinates are used for the 30winter groups, and the following 

 winters respectively, a dot being placed at each crossing point 

 of the lines from the two scales. Alex. B. MacDowall. 



