76 



NA TURE 



[June 22, 1893 



stronger. Thus, excluding insignificant decimals from the sun- 

 spot figures, and taking the mean of three and a half cycles for 

 the Ceylon dates from 1854 to 1S9.1, and from 1864 to 1885 for 

 the sunspots in pairs of years from Wolf's tables (the only sun- 

 spot data I have available) we get the following comparison : — 



Ceylon 1 monsoon dates. Sunspots. 



Mean abnormals. Mean abnormals. 

 1867-78. 



1856 

 1857 

 1858 

 1859 



i860 

 I86I 

 1862 

 1863 

 1864 

 1865 

 1866 



Better sunspot data would certainly not invalidate the connec- 

 tion. The lag behind the maximum sunspot data and the appa- 

 rent tendency to precede the minimum has always been noticed 

 in other phenomena. Moreover, from the analogy between the 

 abnormals of the two elements compared both in quantity as 

 well as sign the same remarks as to the reality of the cycle 

 made by Mr. Blanford in his work (cited ante) p. 254 apply 

 pari passu to that in the Ceylon dates. 



A similar relation holds good for the little monsoon which 

 may be put into words as early dates in years with increasing 

 sunspot numbers and late dates in years with diminishing sun- 

 spot numbers, with a decided maximum of twelve days early in 

 the year immediately succeeding that of maximum sunspots. 

 Even the period between the two bursts shows symptoms of a 

 similar relation to the sun's condition, the mean maximum 

 interval, forty-three days, corresponding to the year of minimum 

 sunspots, and the minimum twenty days occurring two yeais 

 after that of maximum sunspots. The relation, however, is 

 clearest in the figures for the burst of the big monsoon and 

 seems to show that apart from all indirect inttuences such as 

 accumulation of snows on the Himalayan outer ranges, and 

 unusual winter rainfall on the plains or the reverse, there is a 

 real fluctuation in the dates of the burst of the big monsoon or 

 burra barsat connected with the sun's condition which appears 

 to be more direct than that exhibited by the amount of rain 

 which falls during its continuance and appears to indicate, as 

 indeed is borne out by what we know from other sources, that 

 in years of many spots the conditions which usher in the summer 

 monsoon rains are earlier developed, and, as the amounts show, 

 probably continue more regularly than in years of few spots. 



Granting this as a working hypothesis two important results 

 follow. 



(1) The parallel march of the Ceylon dates and the rainfall 

 of the Carnatic shows that the former could be employed to 

 forecast the probable amount of monsoon rainfall about to be 

 enjoyed in the latter district. 



(2) That by using the mean abnormal of the year in its posi- 

 tion in the sunspot cycle as the true mean instead of the mean 

 of the whole period, the true abnormal for the year can be better 

 estimated and the probable general character of the weather 

 foretold. 



As an example let us take the well-known diurnal variation of 

 barometric pressure, whose amplitude in the tropics is so large 

 that it bears a sensible ratio to the abnormal fluctuation pro- 

 duced by a passing disturbance. 



In estimating the true abnormal at some particular hour of 

 the day we must evidently compare the value with reference to 

 the normal at that hour. 



Similarly lor the sunspot period in the case under considera- 

 tion. If there is reason to believe that the period exists we ought 

 to treat it as a reality, and in constructing graphic abnormals 

 take the curve of the progressive cyclic normal as our abscissa 

 axis instead of a straight line representing an endless repetition 

 of the mean of the whole period. The principle is adopted as 

 regards varying locality in drawing synoptic abnormal charts. 

 It should be equally imperative in cases where the element of 

 time is considered. 



Thus in 1894, if the monsoon burst in Colombo twelve days 

 before its time it would be abnormal to the extent of -f 2. On 

 the other hand, if it were twelve days late, it would be abnormal 



1 These figures are simple means itnsuwothed. 



NO. 1234, VOL. 48] 



to the existing mean to the extent of — 22, and even to the new 

 mean formed by incorporating this fresh value, to the extent of 

 - 16, and we might in such a case infer that some unusual cause 

 was in operation which would certainly bode ill for the Madri* 

 agriculturists. 



I have put these facts and considerations forward simply as a 

 preliminary inspection of two phenomena which not only occur 

 in Ceylon, but are more or less common to the Indian peninsula, 

 and to show how conditions, the relations between which can at 

 present only be exhibited in an empirical form, may yet be 

 employed as a means of forecasting the character of a season, 

 and also ultimately by further investigation help to elucidate 

 the whole machinery by which the grand weather changes are 

 produced by terrestrial physical conditions in conjunction with 

 alterations in the state of the sun's surface as well as its varying 

 declination. A large field on either side of the equator, 

 embracing one-fourth of the entire area of the world, exists, 

 from which observations are very much wanted to complete 

 our knowledge of the causes of phenomena which, while 

 they are evidently closely related to action-centres (using 

 Teisserenc de Bort's significant expression) at some distance 

 from the equator, are yet, probably to some considerable 

 extent, dependent upon conditions prevailing over the entire 

 equatorial belt, which may, for all we know, fluctuate in 

 stricter unison with solar changes than those which occur io 

 higher latitudes. E. Douglas Archibald. 



Singular Swarms of Flies. 



With the writer's permission I send you herewith a letter 

 which I have received concerning the subject of my letter which 

 appeared in your issue of June i. 



Daring the week following the date of my letter I repeatedly 

 saw swarms of similar kind ; but smaller and less marked,;] 

 seldom visible much more than fifty yards away ; always undei 

 similar atmospheric conditions, which were chronic during th^ 

 period in question. The swarms always showed much the 

 same slant from vertical (some 30° or so), the direction of ih^ 

 slope in plan being towards such slight draft of air as waffl 

 perceptible. R. E. Fro JDE. j 



Gosport, June 12. 



I FIND in Nature, June i, an inquiry you make abcut flies: 

 forming clouds, resemblmg smoke. 



They are usually produced by the gnats called sclenlificall) 

 Chiroiiomus, and have been often mentioned in entomologic 

 literature. 



I give below several references I can lay my hand on, but' 

 there are probably many more recent ones, which I have not,, 

 noticed — ,i; 



German, Magazin fiir Entomologie (in German), vol. i. pk| 

 134-140, 1813. 



Clapton, J. C. Dale, in Magaz. Nat. Hist., 1833, p. 544,- 

 (In Ireland and England.) <l 



Patterson, Ann. and Mag. of Nat. Hist., vol. x. 1842, p; 

 6-9. 



I have seen such clouds myself more than once. Cases have i 

 occurred when the smoke-like appearance has caused a fire alarm 

 to be sounded. C. R. Osten Sacke.n. , 



Heidelberg, Germany, June 4. 



OFFICIAL CATALOGUE OF THE EXH/B!-\ 

 TION OF THE GE R.HAN EMPIRE AT THE \ 

 COLUMBIAN UNIVERSAL EXHIBITION IN \ 

 CHICAGO. \ 



GERMANY, not unmindful that America is her best ; 

 customer, will be worthily represented at Chicago. I 

 An elaborate catalogue, in the German language, has j 

 already appeared, and an English translation will shortly | 

 be published. We have been favoured with an advance l 

 copy of the latter, which is by no means amereenumera- \ 

 tion of exhibits. It contains a general introduction, and • 

 a number of original articles by leading experts, " in- 1 

 tended to supply for each department a concisely descrip- ' 

 five survey of its development and present condition." j 

 There is also, in German and English, a special Guide j 

 to the collective exhibition of the German chemical. 



