Observations among the Alps. 313 



and a half from the bulb the tube is widened, so that the following 

 inch and a half may represent the range from 45° F. to 115°F. The 

 tube then finishes in a spheroidal chamber, of which the diameters 

 are about an inch and half an inch. The widened portion of the tube 

 may be dispensed with, as the correction which it serves to ascertain 

 may be otherwise found by means of a Table experimentally con- 

 structed for each instrument. In that case the spheroidal chamber, 

 in which the tube will then terminate at eleven and a half inches 

 from the bulb, should be made somewhat larger. The fluid em- 

 ployed is alcohol coloured with a drop of pure aniline-blue. A con- 

 siderable quantity of air is left in the chamber. As a running column 

 has to be read at a particular instant, great plainness is the first requi- 

 site for the scale. On this account graduation on the tube has not 

 been adopted ; but at an inch and a half from the bulb is attached 

 an ivory scale, nine-tenths of an inch broad and eleven and a half 

 inches long (or somewhat less if the widened tube be dispensed with), 

 its other extremity coinciding with the commencement of the sphe- 

 roidal chamber. This scale is graduated throughout in millimetres. 

 The number of millimetres corresponding to each degree Fahr. on the 

 tube of narrow bore, and to every filth degree from 45° to 115° on 

 the widened tube, should be noted on the back of the scale. 



The principle of the instrument is the same as that of Sir J. Her- 

 schel's ; and it is to be worked according to the directions given by 

 him in 'The Manual of Scientific Enquiry.' It was devised for 

 mountain use, where the weight of the Herschel and the fragility of 

 its internal thermometer are elements of difficulty. It has also the 

 advantage of being less costly. The air-chamber is made to serve the 

 purpose of the screw in the Herschel, viz. that of altering at will, 

 according to circumstances, the range of the thermometer. This is 

 effected by throwing off into the chamber a greater or less quantity 

 of fluid, retaining it there by holding the instrument with the cham- 

 ber end somewhat lower than the bulb, and working with the remain- 

 ing column. As alcohol expands unequally between its freezing- and 

 boiling-points, a small correction is necessary, depending on the tem- 

 perature of the alcohol at the time of working. This temperature is 

 ascertained by noting the point in the widened tube, at which the 

 column stands, when the fluid is thrown off into the chamber. The 

 excess of this temperature above 45° F., the point from which the 

 fluid is thrown orf, has to be added to the temperature between 40° 

 and 45° shown by the head of the working column, in order to have 

 the true temperature. From the openness of the scale, and con- 

 sequent small range of the instrument for any one adjustment, it is 

 necessary to select for working a temperature not much removed from 

 that at which the rise in the sun is equal to the fall in the shade. 

 This temperature, which may be called the temperature of equili- 

 brium, will vary practically, according to the solar intensity, from 

 some 5° F. to 20° F. above the temperature of the surroundiug in- 

 fluences. By driving the fluid into the chamber until the tempera- 

 ture of equilibrium is represented at a point near the middle of the 

 tube, the readings will go on for a considerable time without altering 



