448 



NATURE 



{April \, 1878 



Cumulative Temperatures 



In reference to my letter upon the above subject, whicb was 

 published in your columns of February 21 last, I have received 

 from Prof, de CandoUe, of Geneva, a communication dated 

 March il, in which he calls attention to the fact that in his 

 " Geographic Botanique raisonne," which was published as far 

 back as the year 1855, he recorded the suggestion (made by him- 

 self some ten years previously) of the employment of an uncom- 

 pensated pendulum fitted with a suitable registering apparatus 

 for the determination of cumulative temperatures in connection 

 with the application of meteorology to agriculture and to the 

 geography of plants. 



In the above work (vol. i. pp. 58 and 59) the following pas- 

 sage occurs : — 



*' Les chiffres les plus importants ^ connattre pour les applica- 

 tions de la meteorologie k I'agriculture et a la geographie bo- 

 tanique sont, pour chaque localite, les sommes de temperature 

 au dessus de + i°i + 2°, de + 3°, etc., par annee, saison, mois 

 ou fraction de mois. 



" Pourrait-on obtenir ces valeurs directement par un instru- 

 ment special, qui dispenserait de recourir a des calculs com- 

 pliques, souvent impracticables, dans le systeme actuel des 

 observations meteorologiques? C'est une question que je 

 souniets aux physiciens. Elle m'a preoccupe depuis longtemps, 

 mais je suis loin de posseder les connaissances theoretiques et 

 pratiques necessaires pour arrivir a une solution. J'entrevois la 

 possibilite de construire deux sortes d'instruments qui repon- 

 draient aux conditions desirees ; je les mentionne sans pouvoir 

 indiquer les details d'execution. 



" L'un de ces instruments serait la pendule-thermometre de 

 M. Edmond Becquerel, modifie de telle sorte que les battements 

 par une temperature inferieure a 0°, ou ceux inferieures a -f- 1°, 

 a -h 2°, etc., ne seraient pas comptes." 



"Un autre systeme serait celui de thermometrographes mai-- 

 quant les temperatures superieures 4 tel ou tel degre, et seule- 

 ment celles-]a." 



To this the following foot-note was added : — " II y a plus de 

 dix ans je fit des demarches aupres de deux astronomes, M. 

 Gautier, k Geneve, et M. Arago, a Paris, pour appliquer la 

 pendule a la mesure des temperatures. Je proposals une pendule 

 aussi dilatable que possible sous Taction de la temperature et un 

 ^ompteur adapte a I'instrument. . . . Les honorables savants 

 ..uxquels je m'etais adresse penserent qu'il serait trop difficile de 

 soustraire I'instrument a diverses causes d'erreurs." 



From the above, which was written twenty-three years ago, it 

 is clear that to the eminent botanist must be accorded the merit 

 of priority not only of the suggestion but also of the publication 

 of the idea of the method of averaging temperatures by observa- 

 tions of the pendulum, while to Mr. Stanley must be given the 

 credit of embodying that idea in a practical form and constructing 

 an instrument based upon the principle. 



St. Leonards-on-Sea, March 16 Conrad W. Cooke 



The Wasp and the Spider 



I HAD anticipated in my own mind Mrs. Hubbard's suggestion, 

 and only the great pressure on your space prevented my meeting it 

 in my previous letter. In the first place, my recollection is that 

 the spider was of a kind that spins no web ; like our own grey 

 hunting spider, familiar in the summer on walls and palings. In 

 the next place no species of spider, except the gossamers, 

 habitually leaves this fine line behind it. It is in all cases a 

 voluntary act, preceded by a perceptible pause, and pressure 

 downwards of the extremity of the body to attach the end, 

 whether for suspension, or in the process of forming the web. 

 Even the gossamers are no exception to this rule ; only in their 

 case the line, in summer and autumn, is more continuously run 

 out as a point of departure for their mysterious aerial flights. A 

 house-spider, for instance, as he runs across the floor or across 

 your hand, leaves no fine line behind him. The tiny gossamer 

 has an amazing command of the material, but in the larger, web- 

 spinning kinds it is far from inexhaustible, and, at all events, an 

 apparently useless waste is not in the ordinary economy of 

 nature. Moreover, in the case in question the spider was keenly 

 aware he was pursued, and would not willingly leave so fatal a 

 clue on his track. Mr. Merlin, who is on the list for 187833 

 our consul for the Piraeus, is, however, a competent observer, 

 and could settle the question, Henry Cecil 



Bregner, Bournemouth, March 23 



SUN-SPOTS AND RAINFALL 



■D Y the overland mail which arrived here on January 

 ^ 12, I received, through the courtesy of Dr. W. W. 

 Hunter, two copies of a pamphlet on " The Cycle of 

 Drought and Famine in Southern India," a copy of the 

 Nineteenth Century for November, and a copy of a letter 

 on " The Rainfall in the Temperate Zone in Connection 

 with the Sunspot Cycle," published in Nature (vol. xvii. 



P- 59)- . 



Having previously read notices of the pamphlet and 

 being desirous to see it, I requested its author to favour 

 me with a copy. His rainfall cycle for Madras was, so 

 far as I could learn from newspaper reports, identical 

 with a cycle which I had discovered long before. In 

 my official report for 1875, which was printed and cir- 

 culated in 1876, I gave a resume of the results at which I 

 had arrived from 1872 down to the close of 1875, ^.nd 

 stated that an examination of returns from 144 stations in 

 different parts of the world, as well as of the variations in 

 the levels of European rivers, had led me to the con- 

 clusion that there was a rainfall cycle of the same dura- 

 tion as the sunspot cycle and nearly coincident with it, 

 both the sunspots and the rainfall attaining a minimum 

 in the eleventh, first, and second years of the cycle, and 

 a maximum in the fifth year. Hence when I learned from 

 an abstract of Dr. Hunter's results for Madras that in his 

 "cycle of eleven years both the sunspots and the rainfall 

 reach their minimum in the group consisting of the 

 eleventh, first, and second years, and that both the rain- 

 fall and the sunspots there increase till they both reach 

 their maximum in the fifth year," I was curious to know 

 how his cycle had been made out ; for although I had not 

 the Madras rainfall for each year from 1813 to 1872, yet 

 from the falls in the years of maximum and minimum 

 sunspots which I got in the Proceedings and Transactions 

 of the Institution of Civil Engineers (vol. xxxii.), I in- 

 ferred that the Madras rainfall was not quite so favour- 

 able to my hypothesis as the rainfalls of some other 

 places. As, however, I might be wrong, I applied for a 

 complete table of the Madras rainfall, but without 

 success. 



A remarkable rainfall cycle for Bombay, nearly coin- 

 cident with the sun-spot cycle, had been previously ascer- 

 tained, and a similar cycle, though not so well marked, 

 had also been found by comparing the yearly mean rain- 

 falls of Anjarakandy, Bombay, Calcutta, and Nagpur 

 with Wolf's relative sun-spot numbers. 



I have now the whole history of the Madras cycle 

 before me. The author of the pamphlet says that after 

 many experiments he hit upon a method of working out 

 a cycle. This method consisted in commencing with 

 1876, taking backwards, as far as the register extended, 

 periods of eleven years each, and then finding the mean 

 rainfall for each series of years in the common period. 



The results obtained for Madras by this method are to 

 a considerable extent in conformity with those which I 

 had found for different countries ; but there are discre- 

 pancies, one of the most remarkable of which is that the 

 rainfall in the second year of Dr. Hunter's cycle is greater, 

 instead of less, than the mean rainfall. Still there is a 

 certain amount of coincidence. But as the method used 

 by Dr. Hunter — and I would call special attention to this 

 point — is different from the one by which I found my 

 cycle, his results and mine are not comparable. 



The sun-spot cycle being one of about eleven years, 

 and the maximum epoch occurring, on an average, 37 

 years after the previous minimum, and the next minimum 

 7*4 years after the maximum, I found by experience that 

 the best way of comparing the rainfall and the sun-spots 

 was to start either from a maximum or a minimum year, 

 and then to take the proper number of years before and 

 after the epochal year. Commencing with a maximum 

 year, for instance, I took five years before it and seven 



