PROFESSOR FORBES ON THE TEMPERATURE OF THE EARTH. 193 



whole was entrusted to Mi* Adie, to whose experience and skill I am greatly- 

 indebted for the completion of the whole without any material accident. 



For the larger thermometers it is of great consequence that as little liquid 

 as possible should be contained in the stem of the instruments, otherwise the 

 apparent expansion of the column will depend greatly upon the variable tem- 

 perature of the different parts of the stem, as well as on that of the level. This 

 correction being diflBcult to apply with mathematical exactness, it was desirable 

 to make it as small as possible (although it would be unwise to overlook the 

 correction altogether, as most observers have done). I accordingly had twelve 

 tubes drawn, each of about 26 feet in length ; of the external thickness nearly of 

 a common barometer tube (about half-an-inch), but whose internal diameter 

 was nearly capillary. These were carefully examined throughout, by means of a 

 column of mercui-y passed through them. 



The proportional numbers, representing the calibre of the tubes, were en- 

 tered in a table now before me, con-esponding to every foot of their length ; and 

 the tubes were numbered, so that the degree of uniformity of any portion could 

 at any time be ascertained. From these tubes twelve lengths were cut from the 

 most uniform parts, amounting altogether to about 144 feet, for the construction 

 of the three sets of thermometers. 



It is to be understood that the capillary tube now spoken of was made with 

 a view to reach the surface of the ground, above which the tube should expand 

 into one having degrees of a convenient length. So small, indeed, was the stem 

 compared to the bulb, that a degree of Fahrenheit in the capillary tube would 

 have occupied, in one case, a space not less than 51 inches long. The wide tube 

 to which the scale was attached, had a bore of about tV inch, and was made long 

 enough to include the expected range of temperature at their respective depths. 

 These ranges were, however, in some cases rather under-rated. 



The bulbs were cylindrical, and varied in size from about 6 to 8 inches long, 

 and 1^ or 2^ wide. They were blown at the glass-house separately from the 

 tubes. The deepest thermometers had the largest bulbs and longest degrees, be- 

 cause the required range was less. 



From the length and fineness of the tubes much trouble would have been 

 experienced in filling the thermometers in the usual way. The lower end of the 

 cylindric bulb was, therefore, drawn out into a tube, by which the liquid (freshly 

 boiled alcohol, slightly coloured) was admitted, and it was drawn in by the action 

 of a syringe fixed at the extremity of the long stem. Both ends were then closed 

 in the usual way, an expansion being left at the top as in common alcohol ther- 

 mometers, but most necessary in this case, in order to allow for the changes of 

 temperature to which the instruments were exposed before sinking them in the 

 ground. 



The graduation was one of the most delicate parts of the construction. The 



VOL. XVI. PART II. 3 c 



