226 



THE CIVIL ENGINEER AND ARCHITEtT'S JOURNAL. 



[July, 



METEOROLOGY. 



On Ike Instruments employed in Meteorological Observations. 

 % John Drkw, Esq., F.R.A.S. 



'I'liK noble mansion of Ilartwell, situated in the fertile vale of 

 Aylesbury, and at the foot of the C'hiltern Hills — the stewardship 

 of* which is accepted by those who are anxious to resign one of 

 t;reater responsibility — is not devoid of interest from its historical 

 associations. Here in seclusion resided the last of the Rourbons 

 who bore the title of '" King of France." The visitor to Versailles 

 will call to mind the "Jardin a la Hartwell," which bears testimony 

 to the agreeable recollections of this temporary retreat that followed 

 the recluse when the sequence of events had placed a crown on bis 

 head. In the spacious and elegant library, Louis XVII I. attached 

 his signature to the document which restored him to the throne of 

 liis ancestors. \\'ith objects far other than political, a few lovers 

 of science had assembled in this room, at the invitation of the 

 present proprietor, Dr. Lee, on the -tth of April, 1850, for the 

 purpose of taking into consideration the present state of meteoro- 

 logy, and of adopting such measures as might conduce to its 

 advancement. The result of their deliberations was the formation 

 of a society, to be called the "British Meteorological Society," of 

 which Dr Lee was appointed Treasurer, and James Glaisher, Esq., 

 F.R.S., F.R.A.S. (of the Royal Observatory, Greenwich), Secretary. 

 The second meeting of the Council was held on the 7th of May. 

 AVithin a month the Society has numbered ninety-five meniliers, 

 has elected as President, S. C. Whitbread, Esq., F.R.A.S., as ^■i^;e- 

 Presidents Lord Robert Grosvenor, M.P., Hastings Russell, Esq., 

 M.P., General Sir Thomas Brisbane, K.C.B., F.R.S., and Luke 

 Howard, Esq., F.R.S. 



Trusting that this movement will give an impulse to the studv^ 

 of atmospheric phenomena, and an.tious, as one of the original 

 members of the Council, to promote its objects, I avail myself of 

 your offer to discuss in a popular manner the construction of those 

 instruments employed in meteorological observations, the value of 

 the observations themselves, and other points which may he likely 

 to interest and draw attention to a science which is yet in its 

 infancy — a science which calls upon all who imbibe the vital air, 

 the lumen spirahile colli, to co-operate in ascertaining the effects 

 it produces on our sanitary condition: its powers as the medium 

 of conveying the deadly pestilence or the health-inspiring anti- 

 dote. 



To promote this desirable object, it is my intention to describe 

 the principles of construction of such instruments, and those only, 

 as may assist a labourer in this ample field in contributing his share 

 to the accumulation of phenomena from which we may fairly hope 

 will be eventually deduced the laws regulating the atmosphere — 

 the source of life and support to every member of the human 

 family. 



I. — The Barometer. 



Fill a glass tube 32 inches in length with mercury, invert it in a 

 vessel of the same liquid, and you have a barometer: the column 

 of mercury in the tube will descend until its weight exactly balances 

 the pressure of the atmosphere on the surface which is open to its 

 influence. A scale, measuring the height of the top of the column 

 above the surface of the liquid in the vessel, will show from time to 

 time the variations in the pressure of the air vertical to the place 

 of observation. 



Simple as this may appear, w-e find, when we wish to arrive at 

 results sufficiently accurate to be of any value for scientific pur- 

 poses, various sources of error, which must be detected and their 

 value ascertained. The vacuum at the upper part of the tube, if 

 the instrument is well constructed, is the most perfect that can be 

 produced ; to i)rcvent the rise of particles of air whicli may be 

 diffused throughout the mercury, or may have been attached to 

 tlie sides of the tube, the mercury should be boiled in the tube, 

 and the perfection of the vacuum may be tested by inclining the 

 tube and driving the mercury to the closed end, on striking which 

 it «ill give a sharp and sudden tap if no air or moisture exist above 

 the mercurial column. 



If a piece of glass tube, not more, we will suppose, than •l-inch 

 in diameter, be inserted in water, the water will rise within it 

 by capillary attraction to a height greater or less according to the 

 size of the tube — the surface of the water within being concave : 

 on the contrary, if the same tube be plunged into mercury, it will 

 repel the metal all around, and the surface of the mercury within 

 tlie tube will be convex, the top of the curve being depressed below 

 the level of the liquid in the vessel. Now, unless the tube of the 



barometer is so large* that the the capillary attraction may be 

 disregarded, it is evident that a correction must be applied to the 

 observed height of the mercury in the barometer to reduce it to the 

 true. This correction is alviays -)-, and is usually determined by 

 the maker : if it be not, it may readily be obtained from tables 

 when the diameter of the tube is known. 



The scales of barometers adapted to scientific use are of brass 

 throughout, extending from the cistern to the top of the tube : an 

 increase of heat will be followed by an expansion both of tlie mercury 

 and the scale. If the two metals ex])anded equally for equal incre- 

 ments of heat, no error would arise; but mercury expands more than 

 any oth.r metal known. Now supposing the atmospheric pressure to 

 remain the same, but that the temperature has risen within a given 

 ]ieriod from 40° to 60°, the index would show (at a height of about 

 30 inches) arise of 0'054 inch, which would be due not to increased 

 pressui'e, but to the excess of the expansion of the mercury over that 

 of the brass scale. It has been agreed to reduce all observations to 

 a standard temperature, viz., 32 degrees of Fahrenheit — the freez- 

 ing point of water^and for this purpose corrections are tabulated, 

 and may be obtained by inspection. 



In most barometers a thermometer with its bulb in the cistern 

 shows the temperature of the mercury, and it is ]iresumed that 

 this is the same throughout the column. Sir John Herschelt 

 objects to this arrangement on the gi'ound that it does not give the 

 mean temperature of the whole mass, including the column; but in 

 a room wliere the temperature is not subject to sudden changes, it 

 may fairly be assumed that no error will arise from this source. 



When the atmospheric pressure diminishes, the mercury sinks 

 in the tube, and consequently causes a rise of the surface of that in 

 the cistern. The height measured by the scale, supposing it to be 

 fixed, will not then be the true, as its divisions presume the level 

 of the surface to be constant and not fluctuating — in fact, there 

 will only be one point at which the measured distance will 

 exactly agree with the real distance of the top of the column from 

 the surface of the mercury in the cistern. This is termed the 

 neutral point, and is ascertained experimentally by the maker 

 during the ])rogress of construction, and engraved on the scale, 

 together with the proportion between the area of a section of the 

 tube and a section of the cistern. It is evident that the surface of 

 the mercury in the cistern will be lower tlian the zero point of the 

 scale when the reading is above the neutral point, from the abstrac- 

 tion of a portion of its contents to supply the rise in the tube ; and 

 that it will be higher when the reading is below the neutral point. 

 If the capacities be as one to forty-two, one forty-second part of 

 the difference between the neutral point and any particular reading 

 must be added in the former case, and subtracted in the latter, to 

 obtain a corrected height. 



Mr. Glaisher is not friendly to barometers to which it is necessary 

 to apply the capacity correction. The glass tube, having a certain 

 thickness, dips, of course, into the cistern: as the cistern gets 

 fuller fi'om a decrease in the atmospheric pressure, the mercury 

 encloses a greater portion of this hollow cylinder of glass, and 

 therefore rises higher (i. e. the surface approaches nearer the top 

 of that in tube) than is shown by the capacity correction, which, as 

 we have seen, only compares the area of the hollow part of the tube 

 with that of the cistern. Again, vvhen the rise of the mercury in 

 the tube from increased pressure, draws the supply from the cistern, 

 the surface is reduced in height by a quantity dependent on the 

 volume of the section of the tube which it had enclosed, beyond 

 what the correction for capacity would indicate. Nor is it an easy 

 matter to calculate the effect which these .annulus-like sections of 

 the glass tube would produce on the readings of the barometer at 

 different heights. He has, in consequence, superintended the con- 

 struction of barometers in which the capacity correction should be 

 sujierseded, at a very reduced price; and, as it is of importance that 

 all observers who communicate with him should be possessed of 

 instruments which should give uniform results, I shall describe the 

 construction more at length. They are made by Barrow, of 

 Oxenden-street, of whom they may be procured, price 71. 7s., and 

 are all compared with the same standard by Mr. Glaisher himself, 

 who supplies the results with the instrument when it is sent out 

 into the world. 



The cistern consists of a hollow cylinder of glass closed by a 

 leathern bottom. A small index points downward towards tlie 

 surface of the mercury, and the first step in taking an observation 

 is by means of a screw which acts on the leathern extremity of the 

 cistern, to adjust the level of the mercury until it exactly touches 



* The vltanieter of the tube of tlie Greenwich ftandarJ is 0-5U.> iucli ; the correctiuii for 

 caiiai-ily is u-2il2 inch. 

 1 ' Ailiuiralty Manual of Scientific Enquirj.* 



