320 



NA TURE 



[February 2, 1905 



The Moon and the Barometer. 



It is an old popular belief that weather tends to be 

 more settled about full moon. Here are some sayings 

 from Inward's "Weather Lore": — 



" The three days of the change of the moon from the 

 way to the wane we get no rain " (United States). 



" The weather is generally clearer at the full than at 

 the other ages of the moon " (Bacon). 



" In Western Kansas it is said that when the moon is 

 near full it never storms." 



" The full moon brings fine weather." " The full moon 

 eats clouds." (This disappearance of cloud Mr. G. F. 

 Chambers pronounces " a thoroughly well authenticated 

 fact.") 



The following evidence in this connection seems to me 

 instructive. It relates to Ben Nevis (1884-1892, nine 

 years) and Greenwich (1889-1904, sixteen years), and to 

 the summer half only (to be more exact, the six lunations 

 commencing with that which had full moon in April). 



The method was as follows : — In the case of Ben Nevis, 

 fourteen columns were arranged for the fourteen days 

 ending with full moon, and fourteen for those following 

 full moon. Each day with barometer under 252 was re- 



o 



,.,;:> 3 I I -3 ■ ^^ 



/3 /• •? 7 •? 3 I I 3 y 7 9 " li 



presented by a dot in those (graduated) columns ; total 407. 

 The dots in each column were then counted, and the 

 sums obtained were added in groups of three (first to 

 third, second to fourth, third to fifth, and so on). Thus 

 we get the upper curve in the diagram. 



In the case of Greenwich, the method was slightly 

 different (see lower horizontal scale). The columns were 

 for seven days about each of the four phases. For com- 

 parison with the Ben Nevis curve we commence with the 

 first day after new moon. The days here considered were 

 those with barometer under 29 6 inches ; total, 476. 



These two curves seem to tell much the same tale ; fe^v 

 days of low barometer about (just after) full and new 

 moon, many such days about (just after) the quarters. 

 Thus, so far as the summer half in those twenty-one 

 years is concerned, the popular belief would appear to be 

 vindicated. 



To give a fuller idea of the relations, I add a table of 

 the maximum and minimum values (each number is, 

 of course, the sum of three) : — 



Ben Nevis 

 Greenwich 



NO. 1840, VOL. 71] 



It will be seen that the chief maximum is about double 

 the chief minimum in one case, and more than double 

 in the other. 



In a dot-diagram, where each day is represented 

 separately according to its barometer (not merely grouped 

 with others as below a certain limit), the contrast between 

 the phases comes out still more clearly. 



The view here given apparently finds support from 

 various quarters. In the Meteorologische Zeitschrift for 

 1900, p. 421, Herr Bornstein gives a curve of pressure 

 for Berlin (May to August in 1883-1900) which is of 

 similar type to those in the diagram. Fr. Dechevrens 

 informs me that the results above given agree with those 

 of his own observations in China, Constantinople, and 

 Jersey. M. Sainte Claire-Deville found the same variation 

 at Cayenne, in French Guiana. 



With regard to the winter half (October to March), the 

 rt'gime would appear to be somewhat different, but I 

 cannot speak definitely of it at present. 



Whether the facts presented be thought to indicate lunar 

 Influence or not, it may be of interest to watch future 

 weather (in the summer half) from the point of view 

 suggested. Alex. B. M.acDow.nLL. 



Reversal in Influence Machines. 



The method suggested for producing reversal on a 

 Voss or Wimshurst will not be found always trustworthy. 

 ■Atmospheric conditions make a great difference. I have 

 been experimenting for more than a year with the view 

 of finding a solution of the reversal problem, and think I 

 have succeeded in tracing the cause, which is primarily 

 connected with dielectric strain. A Wimshurst with the 

 dischargers beyond sparking distance, working at full 

 speed, will often reverse if the discharge is made by 

 suddenly connecting the terminals, but there is no certainty 

 in producing this effect. I have recently constructed an 

 influence machine akin to the Voss except that the re- 

 plenishment is from the back of the disc. Reversal is still 

 the stumbling block, and must occur with fixed inductors, 

 while no plan for controlling the reversal can be relied 

 upon. I should be happy to give any of your correspon- 

 dents fuller particulars of my experiments if they will 

 communicate with me. Charles E. Benham. 



Colchester, January 14. 



Dates of Publication of Scientific Books. 



May I through your columns suggest to publishers — 

 especially of scientific and mathematical books — to give 

 in their catalogues the dates of publication of their books? 

 As a book often gets out of date very soon, such an 

 addition would greatly help those who have no access to 

 good libraries in selecting books to be purchased. I may 

 say that this is done almost invariably in the catalogues 

 of French and German publishers. To take an instance, 

 the Clarendon Press still includes Price's " Infinitesimal 

 Calculus " in its catalogue. Now, although to one who 

 wants to study the subject in an exhaustive manner the 

 book is very valuable, still, to one who wishes to know 

 the principles only, the book is, to say the least, not 

 worth the big price asked for ; and if the date of publi- 

 cation were mentioned in the catalogue, the purchaser 

 would at any rate know that he was not buying an up to 

 date book. R. P. Paraivpye. 



Fergusson College, Poona, India, January i. 



Super-cooled Rain Drops. 



The letter which appeared in your last issue (p. 295) 

 from Mr. Robinson with reference to this interesting 

 phenomenon reminds me of a similar case which I ob- 

 served in Bournemouth during the winter of 1888, and I 

 described in Nature at the time under the title, " Is Hail 

 Ihus Formed?" (vol. xxxvii., p. 295). 



Cecil Carls-Wilson. 



