228 



THE CIVIL ENGINEER AND ARCHITECrS JOURNAL. 



[JlLY, 



the deposition of moisture was the result — a cause entirely distinct 

 from the one alleged. 



1 h.iTe a building detached from my own residence, which for 

 some short time last Christmas was left without a fire. On my 

 entrance one morning, which happened to be warm after a very 

 cold night, I found the walls covered with moisture: tliey had, in 

 in fact, retained the temperature of the night, and the moisture 

 was due to their not having acquired sufficient heat, in the short 

 space of time elapsed, to exceed the temperature of the dew-point ; 

 moisture, therefore, could not hut settle on them, which disap- 

 peared as soon as they had attained a degree of heat rising 

 :il)Ove it. 



III. — Register Thermometers. 



As it is not my intention to enter upon the theory of atmospheric 

 phenomena, which might perhaps form the subject of a future 

 paper, 1 must proceed to describe another necessary adjunct to a 

 meteorological apparatus, namely — the Register Thermometer. 



Automatic registration of atmospheric phenomena has engaged 

 for some time the attention of scientific men. At the observatory 

 of Brussels, M. Quetelet pointed out to me an elegant contrivance, 

 by which a thermometer was made to record its own variations. It 

 was •uspended on its centre of gravity, so that at the freezing 

 point it should hang perfectly horizontal. At any degree of heat 

 above .32^^ Fahrenheit, or the zero of Reaumur, the expansion of the 

 mercury caused a depression of the end of the instrument the 

 farther from the bulb; whereas, below the freezing point, the 

 metal would retreat towards the bulb, and that |)ortion of the tube 

 would be the heavier. The instrument was connected in an inge- 

 nious manner with a system of levers, one of which moved a black- 

 lead pencil, which inscribed the variations of heat on a sheet of 

 paper connected with clock-work, by which it was advanced equal 

 spaces (about one inch each per hour), a new sheet being supplied 

 every day. These sheets gave the minutest variation in the tem- 

 perature during the twenty-four hours. 



For the last two or three years the magnetical, barometrical, 

 and thermometrical observations have been registered at Greenwich 

 by tlie application of photography. A lamp directs its light to the 

 instrument, wliich light is intercepted by the mercury, and pre- 

 vented from leaving a trace on properly prepared pliotographicpaper 

 placed behind. This paper is wound on a cylinder, which is turned 

 round by clock-work, and its indications form an accurate register 

 of the clianges which may have taken place. Mr. Brook received 

 from Government the sum of 500/. for the skill and labour bestowed 

 on bringing this method of registration to perfection. The mem- 

 bers of the British Association for the advancement of Science will 

 call to mind his papers on the subject, which he has moreover fully 

 discussed in the 'Philosophical Transactions.' From the Greenwich 

 Meterological Observations for 181-7, the following account of its 

 especial application to the dry and wet-bulb thermometer is 

 extracted: — 



" These thermometers are mounted under a shed 10 feet square, 

 standing upon posts 9 feet high, and the centres of the bulbs are 

 4- feet above the ground. The bulbs of the thermometers are very 

 large, being cylinders about 8 inches long, and 0'4 inches internal 

 bore. The fluid is quicksilver. One of the thermometer bulbs is 

 covered (in the usual way) with muslin, which is charged with 

 water by capillary passage along lamp-wicks, connected sometimes 

 with one and sometimes with three cisterns of water. There is a 

 coarse screw-motion for raising or depressing the thermometer- 

 frames, so that each can be placed in such a position with regard 

 to the photographic paper that the temperature shown by the 

 thermometer may be recorded upon a convenient part of the paper. 

 'I'he thermometer-frames are covered by plates having longitudinal 

 apertures, so narrow that any light which may pass through them 

 is completely, or almost completely, intercepted by the broad flat 

 column of quicksilver in the thermometer-stalk. Across these plates 

 - a fine wire is placed at every degree; and at the decades of the 

 degrees, and also at 32^, 52^, and 72°, a coarser wire is placed. A 

 camphine lamp (which has however been lately displaced for gas 

 mixed with the vapour of coal naphtha) is placed near to each 

 thermometer, and its light, condensed by a cylindrical lens whose 

 axis is vertical, shines through the thermometer-stalk above the 

 surface of the quicksilver, and forms a well-defined line of light 

 upon the cylinder of paper which is close to it, parallel to the 

 axis of the cylinder. As the cylinder of paper revolves under this 

 light, it leaves a broad sheet of photographic trace, whose breadth 

 (in the direction of the axis) varies with the varying height of the 

 quicksilver in the thermometer-tube. But the light is intercepted 

 bv the wires placed across the tube at every degree ; and there are, 



therefore, left upon the paper corresponding lines, in which there 

 is no photogenic action." It is found that the application of photo- 

 graphic registration has enabled two observers to record more 

 valuable observations than four were able to do before its intro- 

 duction. 



Private observers, however, cannot be expected to procure the 

 costly apparatus necessary for these elaborate records. The 

 maximum and minimum thermometers invented by Dr. John 

 llutlierford, and described in the 'Edin. Phil. Trans.,' Vol. III., 

 will enable them to record tlie greatest heat during the day, and 

 tlie least during the night, with great certainty and very little 

 trouble. The maximum thermometer is mercurial, and the tube 

 is in a horizontal position; the mercury, as it expands, drives 

 before it an index of steel, which, as the mercury contracts, is left 

 at a point which indicates the greatest degree of heat attained. 

 This thermometer is usually read at 9 a.m., and the index is 

 brought to touch the mercury in preparation for the next day, 

 either by inclining the tube, or by means of a magnet if it does 

 not move freely. In some maximum thermometers a small piece 

 of enamel is introduced between the index and the mercury, to 

 prevent adhesion between the two metals, whereby the index would 

 be drawn back and the observation lost. The minimum thermo- 

 meter is filled with spirits of wine; a small index of ivorj' lies in 

 the spirit, and is drawn backward as the liquid contracts in cooling, 

 for the last film of the column of spirit, from the attraction between 

 it and the interior of the tube, is sufficient to carry back the index 

 towards the bulb: on expansion by heat the spirit, however, freely 

 passes it, and leaves it to point out the lowest temperature attained. 

 These thermometers should be compared with the dry bulb ther- 

 mometers, by immersion in water as before described, and their 

 readings, if they diff'er, registered and applied as corrections. 



It has been usually supposed that a mean of the maximum and 

 minimum readings for a month, divided by the number of days, 

 would give the mean temperature for the month; but one of the 

 contributions of Mr. Glaisher to meteorological science has shown 

 that each month requires a certain quantity to be subtracted to 

 arrive at the true mean. 



He has also shown, that if to the mean of daily observations taken 

 at any hour, certain quantities be added or subtracted, the mean 

 temperature will be the result; it follows, that if the mean tem- 

 perature thus deduced from one or more observations daily, agree 

 with that derived for the corrected maxima and minima, such 

 agreement is a proof of the excellencv and a test of the accuracy 

 of the whole series. (See the 'Phil. Trans.', Part I., 1848.) 



I n illustration I subjoin the result of my observations for January, 

 1849:— 



Mean of the 9 a.m. observations for the month . . •10-7 

 Tabular correction + !• 



Mean temperature for the month. 



Mean of the 3 p M. observations . 

 Tabular correction 



41-7 



441 

 - 2-5 



Mean temperature 41 '6 



Mean of the 9 p.m< 

 Tabular correction 



observations 



Mean temperature 



Mean of the above three results, 415. 



Mean of the maximum readings for the month 

 Mean of the minimum readings . . . 



Arithmetical mean of these quantities 

 Tabular correction 



40-9 

 - -4 



41-3 



451 



38-4 



41-7 

 - -2 



Mean temperature from the maxima and minima 415 

 which exactly agrees with that deduced from the observations at 

 9 A.M., 3 p.M, and 9 p.m., after the application of the proper cor- 

 rections. 



As this paper is devoted to the description of instruments and 

 not to the theory of meteorology, I must forbear to enter further 

 on these points: I have alluded to them sufficiently to show that 

 there is a field of inquiry open for the curious and inquisitive, 

 which will amply repay cultivation. 



Some observers are in the habit of recording the highest reading 

 of a thermometer exposed to the full force of the sun's rays. In 

 this case the instrument is of glass, with the degrees marked on 

 the tube itself, to prevent accumulation of heat, reflection, and 



