1921] on Cloudland Studies 423 



humidity were made at the place of observation, but as a rough guide 

 to the prevailing conditions the vapour pressure values recorded in 

 the vicinity (taken from the official British meteorological records) 

 are noted in the last column. 



From March to August, 1920, and from January to June, 1921, 

 some 200 observations were made, mostly at night, and tabulated. 



Description of Instrument. 



The present form of the instrument is shown in Figs. 6 to 11, 

 Figs. 6 and 7 showing the separate parts, and Fig. 8 a section. The 

 thin hollow metal cylinder A carried the thermoscope cell B, immersed 

 in liquid oxygen, all fitted into an outer 2£-litre silvered vacuum 

 flask 0, the liquid oxygen in which was kept at a sufficient height to 

 ensure that no variations of temperature occurred in the cylinder 

 surrounding the cell. A was made of thin sheet copper, brass or 

 aluminium, closed below by lead about 3 or 4 mm. thick, slightly 

 convex to fit the curve of C ; it easily passed through the flask neck, 

 which was about bh cm. in diameter and 6 cm. long, and reached to 

 within 2 cm. of the shoulder of the flask inside. Steadiness was 

 secured by three light springy prongs of German silver, fitting a 

 short way up into the neck of C, a further support being provided 

 by the tube E leading up from the thermoscope cell and bent over 

 the neck to make a springy contact inside and out. The collar of 

 the base plate D also fitted securely over the flask neck, and had a 

 vertical trough into which the cell exit tube E snugly fitted. The 

 possibility of any displacement of the cell and inner cylinder was 

 thus reduced to a minimum. E was bent out horizontally for the 

 short rubber connection R to the scaled index tube (t in Fig. 10). 

 A sleeve of thin baize, soaked in melted paraffin, was first fitted to 

 the neck of the flask, so that a tight fit of the metal collar and cell 

 tube should be possible without any dangerous pressure on the glass 

 neck. The cell B was made with a cone-jointed ring to carry the 

 rubber membrane, or silver chloride or rock salt plate on a light lead 

 dome, soldered to the coned ring as shown at P, in Fig. 7. The 

 coned ring could thus be manipulated apart from the main body of 

 B, containing the charcoal : the two parts being fitted together 

 warm, with the joint luted by a little freshly melted rubber which 

 rendered it air-tight at low temperatures. Ebonite can be used for 

 the tube E and shutter connecting rod M (Fig. 8), where they pass 

 through the neck of the flask ; screwed connections are then made 

 to the thin German silver tube and shutter hook within the flask. 

 Here also a little melted rubber ensures an air-tight joint in the 

 tube E within the flask. It may be mentioned that a very small 

 leak in the cooled thermoscope is indicated by a persistent high 

 " zero " on the instrument when all is equilibrated, the manometer 

 Vol. XXIII. (No. 115) 2 g 



