4o8 



NATURE 



I Mar. 2 1, 1872 



changes in the temperature of the diy air alone, is the 

 object of the present paper. I commence with the diurnal 

 variation. 



Let us suppose an atmosphere of dry air hardly absorb- 

 ing any heat from the solar rays, and therefore chiefly 

 heated and cooled by contact with the earth. Let us take 

 the moment when the earth first begins to be heated by 

 the sun's rays. (This will probably take place a little be- 

 fore sunrise, in consequence of the large amount of re- 

 flected or diffused heat which accompanies the morning 

 twilight.) The earth then becomes heated at A, while at 

 B, a little more to the west, no heat is yet felt. The earth 



at A communicates its heat to the air in contact with it, aiid 

 the latter expands and becomes lighter than the air in 

 contact with the eaith at B. (At C of course the earth is 

 more highly heated than at A, and therefore the air in 

 contact Sith the earth at C is still lighter.) The imme- 

 diate consequence is that the heavier air at B rushes into 

 the heated space A D (see fig.), driving out the lighter 

 air which occupies it ; and A D becoming filled with 

 heavier air than before, the barometer at A rises. The 

 heating goes on however at A, which, remains at 

 a higher temperature than B, until the epoch of 

 o-reatest heat arrives ; and consequently during all this 

 time there is a flow of air from B towards A next the 

 earth, with a flow in the contrary direction at a 

 greater elevation. It might at first sight appear that the 

 barometer at A would go on rising all this time. But a 

 moment's reflection will show us that though it does so 

 at first, it could not continue to do so all through. For 

 as at the epoch of greatest cold (with which we com- 

 menced) C, A, and B were sensibly at the same tempera- 

 ture, so thev will arrive at sensibly the same temperature 

 at the epoch of greatest heat ; and immediately after- 

 wards the direction of the under-current will be reversed, 

 C having become colder than A, while B is hotter. It 

 is therefore evident that during the whole time which has 

 elapsed between the epochs of greatest cold and greatest 

 heat, the two currents will have counter-balanced each 

 other, the under current having carried exactly as much air 

 from B to A as the upper- current has carried from A to 

 B. Making a somewhat rough approximation, we may 

 assume that during the first half of this period the under- 

 current has been m excess, and the barometer at A has 

 risen, while in the latter half the upper-current has been 

 in excess, and the barometer at A has been falling. Im- 

 mediately after the epoch of greatest heat, the cooler and 

 heavitr air at C will displace the air in the space A D, 

 causing the barometer at A to rise. The motiient of 

 greatest heat will, therefore, correspond to a minimum 

 reading of the barometer, not a maximum ; and after it the 

 barometer will go on rising until half way between it and 

 the moment of greatest cold, when it will again fall until 

 the la-.ter moment. The barometer will, therefore, attain 

 its minimum height at the hours of greatest heat and 

 o-reatest cold, while the maximum heights will occur at 

 about halfway between these epochs. Now this i-esult 

 appears to conform exactly to observation. It must be 

 recollected that the minimum of temperature occurs not 

 more than half an hour before sunrise, while the maxi- 

 mum is generally not reached for two or three hours after 

 noon. This will explain why the morning barometric maxi- 

 mum seems tobenearly an hourearlierthanthe evening one. 

 Indeed observation corresponds so exactly with the re- 

 sults arrived at, that I think h will appear that they cannot 

 be seriously modified by the presence of aqueous vapour. 



The mean of barometric pressures at different latitudes 

 confirms these results. If the trade-winds extended to 

 the poles — which they probably would do were it not that 

 the parallels of latitude become so narrow before reaching 

 them — on the same principles we might expect a minimum 

 of pressure at the equator and the poles with a maximum 

 at alatitudeofabout45^ Forthe second of theseminiina we 

 must evidently substitute the limit of the trades, or rather 

 perhaps of the anti-trades, since the latter seem ultimately 

 to become the under-currents ; and our maximum will 

 be situated about halfway between this limit and the 

 equator. This agrees with observation. The phenomena 

 of the tides too are analogous. There is low water where 

 the moon's attraction is strongest and where it is feeblest, 

 while high water corresponds to the mean attraction. 

 Putting heat for attraction and the sun for the moon, the 

 diurnal variations of the barometer follow the same law. 



This law, however, does not appear to hold so well for 

 the annual barometric changes. We can hardly trace in 

 this case a double maximum in May and November, with 

 minima in January and July. I think, however, that 

 this result may be in part at least explained by the 

 northern and southern shifting of the system of trades 

 and anti-trades. For example, if a place in the northern 

 hemisphere be near this limit (which corresponds to a 

 minimum), the southern movement of the system in 

 winter may cause the barometer to rise instead of falling 

 as we approach the coldest day (supposing of course that 

 it lies to the north of it). On the other hand, at a locality 

 a little to the south of the limit, the northern movement of 

 the system in summer may cause the barometer to rise at 

 the time of greatest heat. I should perhaps notice, how- 

 ever, that the results here arrived at suppose the three 

 points A, B, C to be situated on a horizontal plane, and 

 the specific heat and conductibility of the earth at each of 

 these points to be nearly identical. Hence they cannot 

 be expected to hold for very elevated positions, or for 

 places situated on the sea coast, or the shores of a large 

 lake. They will be found most accurate in the interior of 

 continents, where the land is level, and where the amount 

 of aqueous vapour in the air is comparatively small. This 

 anticipation is also verified by observation, so far as my 

 knowledge reaches. W. H. S. MoNCK 



REMARKS ON THE ADAPTIVE COLOURA- 

 TION OF MOLLUSC A* 



NATURALISTS have long recognised the curious 

 cases oftentimes occurring, of the resemblance 

 between the colour of an animal and its immediate sur- 

 roundings. It had been supposed that climatic influences, 

 or peculiarities of food, or greater or less access to light, 

 had something to do with these coincidences. Mr. Alfred 

 R. Wallace has shown that the varied phases of these 

 phenomena could not be explained by such agents, and in 

 a paper " On Mimicry and other protective resemblances 

 among Animals," published in the Wcstiniiisicr Review, 

 July 1867, and since made widely public in his work on 

 " Natural Selection," he shows that the singular resem- 

 blances between the colour of animals and their surround- 

 ings are mainly brought about by the protection afforded 

 them through greater concealment. Many very interest- 

 ing examples are then cited from the Vertebrates and 

 Articulates in support of these views. Briefly may be 

 mentioned, as examples, the almost universal sand colour 

 of those animals inhabiting desert tracts ; the white colour 

 of those animals living amid perpetual snows ; the re- 

 semblance seen again and again between the colour of 

 many insects and the places they frequent. Among the 

 hosts of examples cited by Mr. Wallace as illustrating 

 plainly the views he advances, may be mentioned the 



* From the Proceedings of the Boston Society of Natur.1l History, vol, 

 xiv., Aprils, 1871. 



