maximum discrepancy of 5 per cent, has been found between Eppley 

 instruments and the new pyrheHometer, and this amount can be reduced 

 while seahng the record. A seven days' record is obtained on a chart 

 placed on the conventional spring-driven drums revolving once a week. 



Discussion 



In reply to a question by Mr. Ewing, Dr. Thomson said that the hydro- 

 gen generating instrument could probably be adapted for use with ferro- 

 silicon instead of aluminium. With regard to the pyrheUometer, Dr. 

 Spilhaus suggested that it would be advantageous if it could be adapted 

 to stand the rolling that takes place on weather ships which at present 

 affects the accuracy of observations taken aboard these vessels. 



With regard to observations in the Arctic, Dr. Priestley raised the 

 question of how one might distinguish between falling and drifting snow^ 

 and in reply Dr. Thomson said that no practical method had been found 

 of making this distinction. 



AUSTRALIAN EXPERIMENTS IN ARTIFICIAL RAIN 

 FORMATION 



By E. BowEN, C.S.I.R., Austraha 



Experiments on the artificial production of rain by the dry-ice process 

 have been conducted by C.S.I.R.'s Division of Radiophysics since 

 January, 1947. The main purpose of the investigation is to obtain a 

 proper understanding of the physical processes which occur when dry 

 icie is dropped into subfreezing clouds. 



The experiments are carried out by C.SJ.R. personnel in aircraft 

 of the Royal Australian Air Force whenever suitable clouds exist in the 

 vicinity of Sydney. These are nearly always found within about 200 

 miles, mostly over, land, but sometimes over the sea. The majority of 

 the tests have been made with cumulus cloud, as this is the type mos:t 

 ■frequently encountered in New South Wales, and it has been found by 

 experience that deep cumulus clouds give the heaviest precipitation. 

 Experiments have also been made with shallow stratus and alto-cumulus 

 cloud, and although precipitation is often induced, in few of these cases 

 has it reached the ground. The quantity of dry ice used has varied, 

 from 50 lb. to 300 lb. dispensed at a rate from 10 lb. to 30 lb. per mile 

 but there is no evidence to show whether different quantities produce 

 different effects. The dry ice is broken into fragments, the majority 

 of which have an effective diameter of about 1 cm. 



Before an experiment is performed, measurements are made of the 

 clear air temperature and humidity in the vicinity of the selected cloud 

 and careful observation made to ensure it does not already contain 

 large drops or ice particles. In some cases flights are made through 

 the cloud before and after seeding with dry ice and observation made 

 of the type of drops and particles encountered and of the vertical air 

 velocity. Visual and photographic observations are made from outside 

 the cloud, and in some cases they are supplemented by 10 cm. and 25 cm. 

 air-borne and ground radar sets. 



128 



