1 5b 



NA TURE 



[June 14, 1894 



puffs were felt very strongly, and especially on the water. At- 

 tempts to measure their temperature with a thermometer proved 

 unavailing, as their temperature was so high and their duration 

 so short that the thermometer had only begun to rise when the 

 heating cause had passed. Between August l8 and 23 exactly 

 the same character of weather was observed in the valleys j 

 around Pontresina in the Engadine. The hot puffs were very ; 

 remarkable. The weather was recognised by the people as 

 "Kihn.'' Here the inadequacy of the thermometer as usually 

 employed was again apparent. Observations were made on the 

 exchange of heat taking place between the hot fiihn wind and 

 the >[orteralsch glacier over which it was blowing. The tem- 

 perature of the air was observed at a station on the glacier, and 

 at a station at the same altitude on the mountain at ils side. 

 The temperature of the air at the land station wa-; found very 

 variable, altering as much as 2' C. in five minutes. By making 

 a number of observations during quarter of an hour a mean value 

 was obtained for comparison with the temperature over the ice. 

 On land the average temperature of the air in the afternoon was 

 l6'5 C. ; over the ice, and at a height of one metre above it the 

 temperature was 10' C, and when the thermometer was laid 

 horizontally with its bulb at distances between two centimetres 

 and two millimetres from the ice, the lowest temperature of the 

 air in that position was 5'''S and the highest 7°"5. The wind 

 blew over the glacier at a speed of eight to ten kilometres per 

 hour, and was a fresh breeze, which might have been expected 

 to thoroughly mix the air, yet the result of repeated observations 

 showed that well-defined temperature gradients were produced 

 and maintained in the air between the ice and a height of a 

 metre alwve it. Between one metre and one millimetre above 

 the ice the gradient is moderate, averaging 3°'5 per metre. In 

 the thin layer next the ice the gradient is precipitous. The 

 occurrence of the highly-heated puffs of air due to the fohn 

 directed attention to the measurement of rapid variations of 

 temperature generally. .\n approximation was made to their 

 estimation by noting the rate at which the thermometer began 

 to rise in one of these puffs, and then determining experiment- 

 ally the excess of temperature of the air required to produce 

 this effect. This could not be satisfactorily done at the time, 

 but attention was paid to it later. 



This method seemed to be the only one by which ordinary 

 thermometers can he made to indicate truthfully changes of 

 temperature which are not extremely slow. The method was 

 applied in the case of a series of temperatures observed at very 

 close intervals during the first two hours after sunrise on sever.al 

 days in February at St. Moritz. The temperature of a ther- 

 mometer freely exposed to the north was observed at intervals 

 of twenty seconds. These are summarised in a table. The 

 temperature exhibited generally two falls for every three to four 

 rises ; the largest rise or fall in twenty seconds was 0°'5 C. 

 From careful experiments on the rate of cooling of the 

 thermometer, both in a room and in the open air when the air 

 was still, it resulted that to produce a rise of temperature ofO'S 

 in twenty seconds, the temperature of the air at the beginning 

 of the interval must have been at least 2°'25 C. higher than 

 that of the thermometer .at the same instant. If there had been 

 a fall of the same amount, then the temperature of the air must 

 have been as much lower than that of the thermometer at the 

 beginning of the interval. -So that if, in any two consecutive 

 intervals of twenty seconds, the thermometer showed a rise of 

 o°'5 and a fall of the same amount, the apparent temperature of 

 the air at the beginning and the end of the interval is the same, 

 whereas the true temperatures differ by 2 x 225 - o '5 or 

 4''oC. A table is given in the paper of the temperatures 

 observed at intervals ol twenty seconds during a few minutes on 

 February 26, when the variations were considerable. The 

 differences of temperatures required to produce the observed 

 changes are given, and from them the amended or true tem- 

 peratures are deduced and tabulated. The true v.ariations of 

 temperature arc naturally much more abrupt than the ap- 

 parent one-s. 



The concluding part of the paper deals with the employment 

 of the thermometer as a caloiimetcr. For this purpose it is 

 necessary, besides the rale of cooling, to know the "thermal 

 mass " or w.iter value of the bulb. A method is indicated by 

 which this can be ascertained with very considerable accuracy 

 by the measurement of the volume of the bulb. The circum- 

 ference of the bulb is liest determinc'l by winding fine thread 

 round it in a close spiral for a certain number of turns, then 

 measuring the length of the unrolled thread. When the specific 



NO. 1285, VOL. 50] 



heat of mercury .and that of ordinary gl.ass are expressed in terms 

 of unit volume they are very nearly identical ; namely, 0449 

 for mercury, and 0466 for glass. It is clear that if the mean 

 of the two is taken to represent the specific heat of the 

 bulb, an error of not more than two per cent, is m,ade in the 

 extreme case when the bulb consists of all mercury or all glass. 

 When the thermal mass and the rateofcoolingof a thermometer 

 have been determined, ils usefulness as a metereological 

 instrument is increased manifold. 



Chemical Society, M.iy 17. — Dr. .Vrmstrong, President, 

 in the chair. — The following papers were read: — The influence 

 of moisture on chemical change, by II. B. Baker. Highly 

 purified lime and carefully dried copper oxide do not combine 

 with sulphur trioxide ; dry ammonium chloride may be sub- 

 limed from a mixture with lime without the liberation of 

 ammonia. Pure dry nitric oxide gives no brown fumes with 

 dry oxygen, although the addition of a trace of moist air causes 

 immediate interaction. Carefully dried ammonium chloride 

 does not dissociate on volatilisation, the vapour having the 

 density 287. — New vol.atile compounds of lead sulphide, by 

 J. B. Hannay. The observed volatility of lead sulphide in 

 water vapour may be explained by assuming the existence of 

 gaseous compounds of lead sulphide and water ; the author 

 concludes that a definite compound of the composition 

 PbS, H.;0 exists. Evidence in support of the existence of a 

 compound of the composition I'bS,SO.j is also brought forward ; 

 both these substances are colourless gases at a red heat, but 

 decompose below 800". — Notes on the cupellation of bismuth- 

 silver alloys, by E. A. Smith. — .Vzo-Z-cresoI derivatives, by 

 R. Meldola and F. Southerden. The authors have endea- 

 voured to determine the constitution of several ortho-.azo-deriva- 

 tives of paracresol by treating their acetyl derivatives with 

 nitric acid or bromine. — Effect of heat on iodates and bro- 

 mates, by E. H. Cook. During the fusion of potassium bro- 

 mate and iodate, bromine and iodine are respectively evolved ; 

 no halogen is given off after melting is complete, and after con- 

 tinued heating to drive off all the oxygen, only potassium 

 bromide or iodide remains. 



Geological Society, May 23. — Dr. Henry Woodward, 

 F.R.S., President, in the chair. — On the stratigraphy and 

 physiography of the Libyan Desert of Egypt, by Captain II. G. 

 Lyons, R.E. The Nubian sandstone, wherever seen, rests un- 

 conformably on the old rocks called by Sir J. W. Dawson 

 Archrean, and the author found no case of alteration of sand- 

 stone by these rock«, though in one case it is altered by an 

 intrusive dolerite. The author considered the Nubian sandstone 

 to be an estuarine deposit which was formed on an area aftcrw.irds 

 gradually invaded by the Cretaceous sea. He considered the 

 whole of the sandstone in the region which he had examined to 

 be of Cretaceous age. He described a series of anticlinals, one 

 set running W.N.W.-E.S.E., and the other N. by E. and S. 

 by W. Many springs of the oases seem to occur along these 

 anticlinals, owing to the beds which contain the water being 

 brought nearer to the surface. Historical evidence was dis- 

 cu.ssed which points to the Nile having reached a higher level 

 in Nubia than it does at present, and it was suggested that 

 variations in the level of the river were caused by earth move- 

 ment opposing obstructions to the river's flow. The sandstone 

 of Jebel Ahinar near Cairo was described, and its occurrence 

 over a wide area west of Cairo recorded. The author con- 

 sidered its age to be later Miocene. He believed that, with the 

 exception of some erosion after the deposition of the Eocene 

 beds, the greatest erosion, including the cutting out of the Nile 

 Valley, look place in Miocene limes, while a certain amount, 

 bringing the area to its pre^ent condition, was done in 

 (Quaternary times. This agrees with the obseivations of the 

 French geologists in Algeria. The origin of the silicificationof 

 the fossil trees of the sandstone-deposits was discussed, and the 

 action of water containing sodium carbon.ate suggested as a cause. 

 The President, Mi. Iludleston, and the Rev, G. Ilenslow 

 having made remarks upon the paper, Prof Hull said he con- 

 curred with the view of the author that the course of the Nile 

 aliovc Cairo had been determined by the line of fault, which 

 follows the valley for many miles upward. As regards the age 

 of the Nile in Egypt, he considered it .as referable to the 

 Miocene stage rather than to the Pliocene. The Miocene 

 period in that part of ihc woild was one in which the main 

 features of the present land-are.as received their general con- 

 tours. Referring to an observation by Mr. Iludleston regard- 





