390 Intelligence and Miscellaneous Articles. 



We know that Young, Arago, and Babinet attributed super- 

 numerary rainbows to effects of diffraction. Airy has developed 

 this theory by mathematical analysis. The results which he ob- 

 tained have been verified by Miller with respect to bows of this 

 kind produced by the sun's rays meeting water-drops from | to 5 

 of a millimetre in diameter. 



According to this theory, the necessary condition for supernu- 

 merary rainbows to be separated so as to be distinct is that the 

 effects of diffraction be produced by a great number of very small 

 and sensibly equal rain-drops. The larger the drops become, the 

 more do the fringes diminish in breadth and the less distinct are 

 they. Thus is explained how it is that in ordinary cases supernu- 

 merary bows are seen only in the upper regions, where the rain- 

 drops are usually smallest — while they do not appear near the hori- 

 zon, where the drops have become larger (Daguin). 



At the time of the observation of Aug. 30 the rain-clouds, 

 moving away from us towards the east, formed two distinct 

 layers : — the one, higher, on which appeared the upper portion of 

 the principal bow, destitute of supernumerary bows ; and the other, 

 appearing nearer to us, extending between the former and the misty 

 zone above mentioned, in front of which the four supernumerary 

 bows that bordered interiorly each extremity of the principal bow 

 seemed to be delineated. Those cloud-arrangements are sometimes 

 exhibited during days of almost continual rain in hilly regions such 

 as that in which I happened to be. 



If it be admitted that the drops of rain proceeding from the se- 

 cond, relatively lower layer of clouds, when near the ground pre- 

 sented still the minuteness necessary to the production of super- 

 numerary bows, their having been seen at the two extremities of the 

 principal bow, near the ground, will be intelligible. If, contrarily 

 to the facts hitherto observed, these bows were not distinct at the 

 upper part of the ordinary bow, it is because the rain-drops which 

 traversed the upper regions of the air, proceeding from clouds much 

 more elevated than those of the second layer, were doubtless large 

 enough to render impossible the effects of diffraction which pro- 

 duce the supernumerary bows, the appearance of which is, more- 

 over, often enough an exception. Under these conditions the 

 effects of the phenomena of refraction and dispersion which pro- 

 duce ordinary rainbows were alone visible in the higher regions of 

 the air. — Extrait cles Bulletins cle V Academic Royale de Belgique, 

 serie 2, t. xlviii. nos. 9 & 10, 1879. 



ON THE DETERMINATION OF HIGH TEMPERATURES. 

 BY H. SAINTE-CLAIRE DEVILLE AND L. TROOST. 



A certain number of chemists are at present engaged in the 

 more or less precise determination of elevated temperatures. At 

 the meeting of the Academy on the 23rd September, 1878, one of 

 us in conjunction with M. Debray announced, in a Note upon the 

 dissociation of the oxides of the platinum group, that the measure- 

 ment of those temperatures had been accomplished by means of the 



