TRANSACTIONS OF SECTION A. 345 



To obtain this value the records of seventy-four balloon nights, some manned, 

 the others free, made on cloudless days, have been assembled and studied. The 

 results show that the amount of water-vapour per unit volume decreases with 

 elevation in an approximate geometric ratio, and that the thickness of the 

 water layer that would result from a condensation of all the water-vapour in 

 the atmosphere above any given level, whatever the season, so long as only 

 cloudless days are considered, may be approximately expressed by the 



equation, 

 x d = 2e, 



in which d is the depth of the water layer in millimetres and e the partial pres- 

 sure of the water-vapour in millimetres of mercury. 



This is some 13 per cent, less than the value found by Hann as the average 

 for all sorts of days, and heretofore used in bolometric work. 



3. Report on the Investigation of the Upper Atmosphere. 

 See Reports, p. 27. 



4. A Theodolite for Observing Balloons. By Dr. W. N. Shaw, F.R.S. 



5. Some Models representing Air-currents up to a Height of Nine Kilo- 

 metres, based upon Observations with Pilot-balloons. By Dr. W. N. 

 Shaw, F.R.S. 



6. Planetary Circulation in the Atmosphere. By Dr. H. N. Dickson. 



WEDNESDAY, SEPTEMBER 6. 



The following Papers and Eeport were read : — 



1. On Possible Relations Between Sun-spots and the Planets. 

 By F. J. M. Stratton. 



The author discussed the sun-spot material for the years 1874-1909 tabulated 

 to show phase effects due to Venus and Jupiter. The general disagreement of 

 the Venue and Jupiter results, when compared with the agreement found 

 between Venus and Mercury by Dr. Schuster, leads to the general conclusion 

 that the inequalities found are to be traced to the chance incidence of the 

 various sun-6pot maxima, and that a considerably longer period must be avail- 

 able before trustworthy results can be obtained. 



2. On the Law3 of Solutions. By H. Davies, B.Sc. 



Ostwald's law does not hold for solutions of binary electrolytes, and empirical 

 relations have been devised by Rudolphi, van't Hofi, Storch, Kohlrausch, &c. The 

 object of this paper is to obtain Rudolphi's and van't Hoff's equations from theoretical 

 considerations and to deduce equations holding for dilute solutions of more complex 

 electrolytes. 



Solvents are characterised by strong association. The cause of the association 

 in the case of water is the fact that oxygen has a potential valency of four. Taking 

 this value, chemical formulae can easily be made up for (H 2 0)n where ' n ' may be 



