TRANSACTIONS OF SECTION A. 51^1 



4. The coronas of the ' polar ' or ' irregular ' type occur about the times wheii 

 the prominences are most abundant near the solar poles. 



5. The ' equatorial ' coronas when there is one centre of prominence action 

 (about latitude ± 45) in each hemisphere. 



6. The •■ intermediate ' type is produced by tico centres of prominence action 

 in each hemisphere, but neither centres near the poles. 



7. The peculiar ' arched ' form of some streamers is produced by the action 

 of two zones of prominences situated near the extremities of their base. 



8. Sun-spot activity has apparently no direct connection with the production 

 of the coronal streamers. 



Report on Meteorological Observations on Ben Nevis. 

 See Reports, p. 56. 



3. Electrical Self-recording Insti'uments. 

 By Professor H. L. Callendar, F.R.S. 



4. Effect of Meteorological Conditions upon Audibility. 

 By A. Lawrence Rotch, B.S., M.A. 



Notwithstanding previous investigations on this subject, the opportunity to 

 determine the variable influence of a stratum of air 600 feet thick, having a 

 meteorological station at the bottom and the Blue Hill Observatory at the top, 

 caused the writer to institute daily observations of audibility during the year 1901. 



The source of sound employed was a steam whistle, distant 2'7 miles on the 

 plain, which was blown twice a day, and the intensity of the sound at the 

 Observatory estimated on a four-part scale. 



The observations were discussed on the hypothesis that variations in the 

 intensity of sound are caused by vertical differences in wind-velocity or in tem- 

 perature and moisture. It is found that differences of temperature and relative 

 humidity between the two stations had no appreciable effect on the audibility, but 

 that the variations observed in it could be explained by the action of winds 

 increasing in velocity with altitude at a known rate, which tilted the sound-wave 

 over the Observatory when the wind was opposed to the sound, and kept it from 

 rising high above the ground when the wind blew from the source of sound. Nearly 

 equal audibility was found for winds blowing at right angles to the above, a 

 phenomenon that was explained by the late Sir G. G. Stokes to the British 

 Association in 1857. 



Measurements of the velocity of the sound, corrected for the temperature of 

 the air, showed an acceleration in winds blowing from the whistle to the observer 

 which approximately equalled the known speed of the air stratum. This investiga- 

 tion will be published in the ' Annals of the Astronomical Observatory of Harvard 

 College,' vol. xliii. Part III. 



5. On some Rainfall Problems. By Hugh Robert Mill, D.Sc, LL.D. 



In attempting to ascertain the distribution of mean rainfall over a large area it is 

 necessary to make allowance for the unequal height of the receiving surface of the 

 rain-gauges above the ground, for the irregular distribution of rain-gauges over the 

 country, and for the different lengths of the records from the various stations. 

 When the object is to ascertain the distribution of rainfall for any given day or 

 month, the hour of reading the rain-gauge and the method of entering the 

 result have to be ascertained, and varying methods adjusted to a common standard. 

 The determination of the distribution of rainfall for a given year involves less 



