TRANSACTIONS OF THE SECTIONS. 31 



situated forest-land, where, at the tops of the trees, ends of branches, and edges of 

 loaves, the tension is high ; and this is in accordance with observation. And 

 secondly, the tension being relatively high at the tops of the elevations of a moun- 

 tainous district, the rainfall should be greater there than in the neighbouring 

 plains ; this, again, is borne out by obsei-vation. Further, at the commencement of 

 a passing thimder-storm, a sudden heavy shower of rain will often faU for a few 

 moments, and then suddenly cease. May not this arise from the approach, by the 

 agency of opposite wind-currents, of detached masses of oppositely charged clouds, 

 the process, just described, of formation of rain-drops going on rapidly in each mass 

 as the two come near each other, and stopping when, by a ilash of lightning between 

 them, the two masses are brought into the same electrical condition ? 



An experimental test of this idea would be to repeat Dalton's measures of the 

 pressure of vapoiu- in the vacuum space of a mercurial barometer-tube (filling that 

 space with air and a little water), and compare the values foimd when the mercury 

 was charged with electricity and when not so charged.! If in the former case a 

 less pressure was found, we might conclude that the particles of vapour are really 

 susceptible of electric induction, and the amount of difference existing would enable 

 us to estimate whether the attractions of the particles upon each other were strong 

 enough to cause the formation of rain-drops hypothetically attributed to them 

 above. 



I 



On a Scale for computing Humidity. By Professor J. D. Evekext, D.O.L. 



The scale in question is the invention of Mr. H. C. Russell, of Sydney Obser- 

 vatory. It consists of a sheet of paper ruled with vertical lines, each corresponding 

 to a degree of the dry-bulb. These are traversed by a set of curved lines, each 

 coiTesponding to a degree of humidity. A detached strip of paper divided into 

 parts which correspond to every even tenth of a degree of difference between dry- 

 and wet-bulb is applied (with its zero on the saturation-line) to the line repre- 

 senting the given diy-bvdb temperature, and the curved line which cuts it at the 

 division corresponding to the given difference indicates the resulting humidity. 



The scale is based on Glaisher's Tables ; and any other table of double entry 

 might be represented by a scale constructed on the same plan. Interpolation is 

 much easier with such a scale than with a table. 



Barometne Predictions of Weather. By Feakcis Galtojt, F.B.8. 



It is notorious that the movements of the barometric columin corrrespond in some 

 sense to the changes of the weather, and especially to those of the wind's velocity ; 

 but they certainly take no notice of the rapid and tumultuous changes of its velo- 

 city which are recorded by the jagged lines of a pressiu-e -anemometer. They there- 

 fore correspond to mean values of the weather ; but the way in which, and the 

 period of time for which those means should be taken, has yet to be determined. 

 Comparison was made between a curve formed on the principle that the ordinate of 

 each point represented the mean velocity of the wind for half an hour pre-vious to, 

 and half an hour subsequent to the moment indicated by the abscissa of that point, 

 and it was di-awn on the same time-scale as the corresponding barogram; the 

 velocity-scale was so adjusted as to allow about the same range in the diagram for 

 the two curves, and the ordinates were measiu-ed from above downwards (were, in 

 fact, negative ordinates), in order that the increase of wind should be indicated by a 

 descending cm-ve, to correspond with the descending barogram, and vice versa. The 

 curve so made was called a curve of l-hom* period av. wind vel., and similar 

 cui'ves were drawn for .3-hom', 6-hour, 12-hour, 16-hour periods, and some others. 

 It was manifest, on comparing these with the barogram, that a period of 1 hour 

 was far too short, for its curve showed many large irregularities, of which the 

 barogram took no cognisance ; a period of 3 hours was much better ; of 6 hours 

 better still ; and the maximum of correspondence began at 12 hours, and ended at 

 16, beyond which time the wind-velocity curve was less irregular than the bare- 



