No. I was a large thermometer, by Adie, at tached to the case of a sympiegometer, of conside rable mass. No. 2, a thermometer by Jones, with a metal scale of the usual size. No. 3, a small and delicate one, with light ivory scale by Cary. These were all hung together. Nos. 1 and 2, by the supe rior conducting power of their scales, kept nearly together till the maximum, but at 3h. 30m. the effect of the mass of No. I becomes evident by keep ing up the temperature, while No. 2, losing its heat with facility, speedily began to decline. Taking also into consideration the third modification of absorp tion, as connected with the colour of the object upon which the instrument is suspended, we pre sent the following table; in which No. 1 is the in strument No. 2 of the last table, but inclosed in a glass cylinder, and hung against a pannel of wood painted black, at the west side of a north window. No. 2 is the thermometer of a sympiesometer, hung on the opposite pannel, exposed without cover to the N.W. wind. No. is the thermometric indica tion of the gaseous bulb of the same instrument, which, as usual, is protected by a polished brass cylinder, but admitting free ventilation. It was used by correcting the instrument for pressure in stead of temperature, as is done when we wish to extract the barometric indications. No. 4 was a small thermometer with detatched scale freely sus pended at the window sill. The differences are suf ficiently marked, but had the day been more sultry and perfectly free from clouds, (which sometimes intervened) they would have been more striking.
tune of the atmosphere by observations at any one spot. Another illustration will be found in Ta ble IV.
IV. I•mturrv.—The first source of error, which we have classed rather from convenience than by strict theory under this head, includes the many irregularities which disturb the indications of the thermometer at the time of dew falling, a subject one of the deepest in meteorology, and which we merely introduced as one of the most important sources of error requiring to be thus scantily men tioned hi this place. And we cannot help observ ing, that in the attempt to register faithfully the true atmospheric temperature for twenty-four hours in summer, the most difficult point of management is about sunset, when we are driven from the trees by the copious deposition of humidity, and tempo rary decline of local heat, to the alternative of a building, the whole of which is more or less affected by the absorbed heat of the preceding day.
In Table XI., No. 1 was exposed in contact with the wall, at a window thirty feet from the ground. No. 2 was suspended four feet above the grass and free from buildings. No. S was laid on the grass and exposed to terrestrial radiation, which, accord ing to the beautiful theory of Dr. Wells, is the ori ginal cause of dew. The difference between Nos. 1 and 3 here amounted to 8°.
The following small table shows the remarkable continuance of the effect of absorption in a large building, especially when accumulated by a black surface, with which the instrument is in immediate contact.
No. 1 is in this situation, being the thermometer with a metallic scale, the same as No. I in the last table, placed, as there described, in a common glass cylinder, on a black pannel facing the east, at a north window. No. 2 is No. 2 of last table, and freely suspended for comparison, the observations commencing above an hour after sunset, and the differences regularly diminishing, and scarcely va nished after an hour and a half, a curious fact, show ing the intrinsic difficulty of obtaining the tempera Table XII. is illustrative of the same facts, and the differences of the temperature of the grass; and the air four feet above it (the numbers being of ex actly the same import as in the last table) are still more remarkable. But we have not room at present to enter into the consideration of these facts.
With regard to the proximity of water, the other source of humidity mentioned, we have nothing particular to add; the cold produced liy the solu tion of water is well known to be subjected to cer tain hygrometric laws, and this influence commu nicates an absolute degree of cold to any film of air actually in contact with water. The proximity of the sea has besides a general influence on climate, by tending to flatten the form of the annual curve, its temperature diverging little from that corres ponding to the latitude.
With these details, for the imperfection of which we have already apologised, we must conclude at present the subject of thermometric errors, and with them the present article. The meagre illus tration we have been able to give of these remarks, show that abundant scope is given for extended researches upon a systematic plan, with the cer tainty of results of practical value, which ere this the author of the present article had hoped to have filled up, but at least we are convinced that we should not have done justice to the thermometer as an instrument, without some attempt to direct the observer in the practical use of it. Our limits do not permit us to enter more fully upon the means of preventing errors in practice, but the detail of the sources of these has almost formed a sufficient guide.
We here subjoin a very complete table of phe nomena occurring at particular temperatures.* For the higher or pyrometric temperatures we refer to the article PynomETEn, as likewise for the expansion of different substances.
On the subject of the THERMOMETER, the follow ing works may be consulted:—Sanctorii Commen laria in ./Ivicennam. Saggi di _Volumle Esperienze, vol. i. Van Helmont Opera; Sturmius, Co&giant Naturate Curiosum, Nuremburg 1666; Martine on Thermometers; Newton, Principia: Hooke, Micro graphia; Boerhaave, Chmire, i. 720; De Luc, Mo difications de P.qtmosphere; Leslie on Heat; Rum ford's Essays; Crawford on .2nimal Heat; Dalton's Essays; Nicol son's Chemistry; enry's Chemistry, vol. i. and Appendix to vol. ii.; Daniell's Meteor ological Essays; Blot, Traite de Physique, vol. i.; Parry's Second Voyage, appendix.