August 1 6, 1877] 



NA TURE 



333 



whole, give higher values for the resistances than those 

 taken when it was small. 



Accordingly, the results were grouped according to the 

 breadth of the black film, and it was found that the mean 

 value of the resistance of a ring i mm. broad, deduced 

 from experiments made when the breadth of the black 

 was between 



These numbers prove that the thickness of the black 

 ring is independent of its breadth. 



It was, however, thought possible that the thickness of 

 the black film might be affected by that of the coloured 

 portion of the film which appeared to be in contact with 

 it, and accordingly the results were once more grouped 

 with reference to that colour onl) . 



The mean value of the resistances when^ the colour 

 next to the black was the 



Blue of the second order, was 



Green ,, ,, 



Yellow ,, ,, 



Orange ,, „ 



Red 



Green of the^ third order ,, 



i-8»6 



1748 

 1719 

 1756 

 I 716 

 1738 



The first of these numbers is considerably larger than 

 the rest, but it is deduced in great part from a high set of 

 measurements which were obtained during the observa- 

 tions on the first cylinder. Equally high values were, 

 however, obtained later on the same day Vvhen the colour 

 nc.\t to the black was the orange of the second order, and 

 on the only other occasion on which an observation was 

 taken with the blue of the second order in contact with 

 the black, the number obtained was 1760. Hence the 

 high value, i '826, does not seem to have any special 

 significance, and we conclude that the thickness of the 

 black is independent of the portion of the film which 

 appears to the naked eye to be in immediate contact 

 with it. 



The principal error with which these experiments are 

 probably affected is due to the fact that the lower boundary 

 of the black part of a film was not always strictly 

 horizontal. When this was noticed to be the case, the 

 breadth of the black ring was measured in several parts, 

 and a mean value was deduced, but it was difficult for 

 the observer to determine whether or no the further edge 

 of the black was below that nearer to him. In spite, 

 however, of this and the other possible errors of 

 experiment, the numbers obtained certainly prove that 

 the thickness of a black film is uniform, and is very 

 approximately constant under such variations of circum- 

 stance as those to which the films observed were exposed. 

 One other point may be worthy of notice. The mean 

 value of the resistance of a black ring i mm. broad 

 deduced from all the experiments was 1,750,000 ohms, 

 and by applying Ohm's law to this we obtain for the 

 thickness of the black film twelve-millionths of a milli- 

 metre. This value, which must be received with the 

 cautions given above, is only one-third of that at which 

 a film of the solution uicd would begin to appear black, 

 and would make the thickness of the film one-forty-ninth 

 part of the wave-length of D. 



As Prof. Reinold and the writer are at present engaged 

 in investigating the question of the magnitude of the 

 radius of molecular attraction, a consideration of the 

 various speculations as tomolecular magnitudes, to which 

 these experiments lead must be for the present postponed. 



A. W. RUCKER 



RAINFALL IN SOUTH INDIA 



THE probability of another failure of rairi and con- 

 sequent famine in South India gives to any facts 

 connected with rainfall in that country so great an interest 

 that I will not delay longer the publication of a result 

 obtained by me several months ago in a paper which 

 would have been presented to the Royal Society last 

 session, but for the desire to complete it with some details 

 expected by me from India. 



Whether the amount of rain follows the decennial law 

 or not, all the known causes of variation will pass, it may 

 be believed, through all their phases within ten years ; so 

 that the yearly mean rainfall deduced fiom ten years' 

 observations should give a considerable approximation to 

 the mean from any series however long. Any deviations 

 from this result would be expected to be small and 

 irregular. The fact is otherwise, if the observations at 

 Trevandrum, on the west coast, and at Madras on the east 

 coast, may be taken (as has been done) for approximate 

 representations of the variations for the country around. 



The Trevandrum series of observations includes the 

 years 1838 to 1876. Taking the sums of rainfall for each 

 ten years (1838-47, 1839 48, and so on), it appears that 

 the amount was a maximum in the ten years 1S43-52, 

 and equal to 761 inches, from which time it has gradually 

 diminished (with occasional slight increases) till now, 

 1867-76, when it is only 562 inches ; or 'Caz yearly mean 

 rainfall during the last ten years was nearly 20 inches 

 less than a quarter of a century ago. 



When we examine the Madras observations we find a 

 quite similar result. The Madras series includes the 

 period 1 8 13 to 1876 : the ten-yearly sum of rain was a 

 maximum in 1818-27, ^"d equal to 555 inches ; it dimin- 

 ished to a minimuin in 1828-37, and equal to 382 inches ; 

 increased, as at Trevandrum, to a maximum in 1843-53, 

 and equal to 583 inches ; diminishing thence to 396 inches 

 in 1S60-69. The agreement with Trevandrum is com- 

 plete till this period ; but the sum increased at Madras 

 to 510 inches in 1865-74, diminished thence till now, 

 whereas at Trevandrum the diminution has continued 

 to the present time. 



If we compare the diminutions at the tvvo stations 

 while they agreed, we find — 



1843-52 



Diminution 



Madras. 



583 inches. 

 396 „ 



1S7 „ 



Trevandium. 

 761 inches. 

 _583 „ 



178 „ 



Or at both stations a diminution of the yearly mean rain- 

 fall of about 18 inches. It will be of importance to 

 ascertain whether this diminution has been experienced 

 at more northern stations. 



The yearly rainfall on the west coast of India, near 

 Cape Comorin, was about 26 inches in 1844 ; it increased 

 to about 70 inches at Trevandrum, thence to about 120 

 inches at Cananore, diminishing at Bombay and farther 

 north, the maximum fall occurring somewhere between 

 Cochin and Bombay. This variation does not depend 

 wholly at least on the nearness of the stations to the 

 range of the Ghats. It appears that the great atmospheric 

 current sweeping over the Indian Ocean charged with 

 vapour during the monsoons has a central current of 

 maximum vapour depth or of velocity. If we can imagine 

 that this central current shifts on the whole northwards 

 from year to year for some time, the stations to the south 

 will receive less and less rain, while those to the north 

 will receive more ; the total precipitation might thus 

 remain the same. If we could suppose such a movement 

 of the vapour masses to obey a law like that of the sun- 

 spots (and something resembling this has been found by 

 me in the investigations for the isobaric lines and wind- 

 currents in the British Isles), then we should find in such 



