Mr. J. J. Waterston on Solar Radiation. 507 



Appendix, describing the Method employed to discover the Influence 

 of the Air in the Cooling of the Sun-thermometer X, and of 

 ascertaining the Correction required to be applied to Observa- 

 tions of r, so as to reduce them to a vacuum. 



The apparatus employed was the vacuum-bath represented in 

 fig. 2. Plate V. (see explanation of figures). With a plentiful 

 supply of lard to the stuffing-box and ground surfaces, a good 

 vacuum could be maintained in it for a day unimpaired. 



The time was measured by the beats of a clock : to register 

 the number of these at each degree as the mercury of the ther- 

 mometer descended, a scale of equal parts was prepared extend- 

 ing to 1000, and with distinguishing marks at each 5, 10, 50, 

 and 100. Then with a pencil in the right hand over the scale, 

 and a magnifying glass in the left over the scale of the thermo- 

 meter, I counted the beats ; and when the mercury came to the 

 line pf a degree, made a mark on the scale of equal parts oppo- 

 site the number of beats, and at the same time continued to 

 count on ; e. g.ii^l was the number when the mercury came to 

 a line, a pencil- mark was made at 5 7 on the scale of equal parts, 

 and the counting went on, 58, 59, &c, until the mercury came 

 to the next line. 



Thus not a beat was lost from beginning to end, and the accu- 

 racy was only limited by the accuracy of the divisions on the 

 scale of the thermometer. Indeed, this method is a severe test 

 to the equality of the divisions, because the reciprocal of the 

 differences in the number of the beats for each degree, if laid 

 off as ordinates to the total number of beats, ought to range in a 

 straight line, and any saw-like irregularities indicate inaccuracy 

 in the divisions of the scale of the thermometer. To heat the 

 bulb of the thermometer, a funnel was placed over the small 

 flame of a Bunsen ; then holding the plate (having the thermo- 

 meter fixed in its place) by means of the stopcock, the bulb was 

 brought over the top of the funnel until the mercury had risen 

 to near the top of the scale. The plate was then quickly placed 

 on the cylinder, communication made with the air-pump, and 

 the air exhausted from the cylinder by twenty strokes, the 

 capacity of the pump being about one-third that of the cylin- 

 drical vessel or vacuum-bath. The vessel thus exhausted was 

 placed in a water-bath, the temperature of which was ascertained 

 at the beginning and end, the difference seldom amounting to 

 one-tenth of a degree. 



The following Table exhibits two series of observations on the 

 cooling of the sun-thermometer X in the vacuum, and in air, 

 taken while the water-bath remained steady at 48°. This basal 

 temperature being at an exact degree, enables the rate of cool- 



