1921] on Cloudland Studies 427 



that the interior was all at a uniform low temperature. In this 

 way the two principal sources of error were eliminated — (1) the 

 entry of stray radiation apart from that direct from the portion of 

 sky to be observed ; and (*2) the fluctuations and contaminations 

 arising from exposure of the instrument in the open air. 



Portable Form. — A Stevenson louvred screen (as used for 

 meteorological instruments) was utilised as a protection equally 

 against bad weather and direct sunshine. By some simple adapta- 

 tions a self-contained and easily portable instrument was evolved. 

 In Fig. 9 one side of the Stevenson screen is seen open to show 

 the thermoscope flask, as already described, fixed in position, stout 

 rubber tubing passing through the slatting to the manometer, which 

 was fixed on a horizontal pivoted batten outside, as shown in the 

 end view (Fig. 10). The three-way cock a connected the cell tube 

 either to the manometer / (horizontal scaled tube) or to the U-tube 

 c, containing a little solid potash or granulated soda lime, and thence 

 by a T-piece/ to the equalising flask d, and also to the other end of 

 t. Inserted through the rubber cork in the neck of d was a thistle 

 funnel e bent over and lightly packed with cotton wool. Gusty 

 fluctuations of the air were thus damped down in entering through e 

 and equally transmitted to each end of the manometer, which thus 

 remained quiet. Unsteadying pulses of pressure back to the cell 

 charcoal were then damped out, and the cell protected from disturb- 

 ance while equilibrating between readings. A thin metal hood (not 

 shown) was screwed over the manometer and fittings to keep them 

 dry ; and in this hood, opposite tbe manometer tube t and stopcock 

 a, a hinged flap was arranged to be hooked up when observing, the 

 sides of the screen remaining closed up. The cell shutter was then 

 raised by turning a milled wheel with a slotted hollow spindle (in a 

 socket in the side of the screen shown open), and engaging the 

 pinion P (Figs. 6 and 8) which actuated the toothed quadrant lever 

 L, for working the connecting rod M, which raised and lowered the 

 shutter S. The details of this will be clear from Figs. 6, 7 and 8. 

 A floor was of course added to the Stevenson screen, and four 

 adjusting screw feet. The instrument was provided with iron 

 handles at each end, and could thus be readily lifted complete and 

 set in any desired position. The ordinary top provided was fitted 

 with a thumb screw at each end, to be easily removable for observa- 

 tions ; the instrument was thus entirely protected, but readily 

 available for use. There was ample space inside the Stevenson 

 screen for a covered gas jar (containing a few pieces of fused calcium 

 chloride under a perforated disc) in which the stopper K and 

 attached bag of uranium nitrate were hung, when removed from the 

 flask. 



Arrangement in Laboratory. — The ordinary form of the instru- 

 ment was arranged on a small gallery erected immediately under a 

 sliding skylight in one of the laboratories of the Institution. A view 



