1885.] on Sunlight and the Earth's Atmosphere. 279 



the surface and at the bottom of that turbid sea where we had left 

 the rest of mankind. We descended the peak and hailed with joy 

 the first arrival of our mule trains with the requisite apparatus at the 

 mountain camp, and found that it had suffered less than might be 

 expected, considering the pathless character of the wilderness. We 

 went to work to build piers and mount telescopes and siderostats, in 

 the scene shown by the next illustration on the screen, taken from a 

 sketch of my own, where these rocks in the immediate foreground 

 rise to thrice the height of St. Paul's. We suffered from cold (the 

 ice forming 3 inches deep in the tents at night) and from mountain 

 sickness, but we were too busy to pay much attention to bodily 

 comfort, and worked with desperate energy to utilise the remaining 

 autumn days, which were all too short. 



Here, as below, the sunlight entered a darkened tent, and was 

 spread into a spectrum, which was explored throughout by the bolo- 

 meter, measuring, on the same separate rays which we had studied 

 below in the desert, all of which were different uj) here, all having 

 grown stronger, but in very different proportions. On the screen is 

 the spectrum as seen in the desert, drawn on a conventional scale, 

 neither prismatic nor normal, but such that the intensity of the energy 

 shall be the same in each part, as it is represented here by these equal 

 perpendiculars in every colour. Fix your attention on these three 

 as types, and you will see better what we found on the mountain, and 

 what we inferred as to the state of things still higher up, at the 

 surface of the aereal sea. 



You will obtain, perhaps, a clearer idea, however, from the follow- 

 ing statement, where I use, not the exact figures used in calculation, 

 but round numbers, to illustrate the process employed. I may premise 

 that the visible spectrum extends from H (in the extreme blue) to A 

 (in the deepest red), or from near 40 (the ray of 40-100, OOOths of a 

 millimetre in wave-length) to near 80. All below 80, to the right, is 

 the invisible infra-red spectrum. 



Now, the shaded curve above the spectrum represents the amount 

 of energy in the sun's rays at the foot of the mountain, and was 

 obtained in this way : — Fix your attention for a moment on any single 

 part of the spectrum, for instance, that whose wave-length is 60. If 

 the heat in this ray, as represented by the bolometer at the foot of 

 the mountain, was (let us suppose j 2°, on any arbitrary scale we draw 

 a vertical line, 2 inches, or 2 feet high over that part of the spectrum. 

 If the heat at another point, such as 40, were but a 4-°, a line would 

 be drawn there a quarter of an inch high, and so on, till these vertical 

 lines mark out the shaded parts of the drawing, the gaps and de- 

 pressions in whose outline correspond to the " cold bands " already 

 spoken of. Again, if on top of the mountain we measure all these 

 over once more, we shall find all are hotter, so that we must up there 

 make all our lines higher, but in very different proportions. At 60, 

 for instance, the heat (and light) may have grown from 2° to 3^, or 

 increased one-half, while above 40 the heat (and light) may have 



