^4) The mean eddy-flux of momentum found for the year at Norfolk 

 Island was 0-16 x 10^ dynes /cm., in such a sense as to exert 

 an eastward pull on the tropical zone. From this result the 

 eddy-flux of momentum, or total Reynolds' stress across 

 latitude 30°, does not appear to maintain the surface easterlies 

 on the equatorial side nor the wide belt of westerlies to pole- 

 ward. For the maintenance of these mean zonal flows we 

 require a far larger poleward flow of westerly momentum 

 across this critical latitude, and this momentum transfer must 

 inhere in the mean flow pattern. A toroidal circulation, with 

 poleward component of flow at the level of maximum wind 

 (jet stream), would supply a transfer in the right sense. The 

 danger of the geostrophic wind assumption in these studies is 

 to be emphasized. 



<5) The total eddy-flows of heat in all these latitudes are of the right 

 sign and magnitude to compensate for the radiation losses and 

 gains. But since^mean toroidal flows, which are necessary for 

 the mass and momentum balance, must bring about a large net 

 flow of heat (see references), it is clear that the mean tempera- 

 ture and moisture distributions are not to be explained in 

 terms of eddy-transport alone. 



References 



Priestley, C. H. B. : U.G.GJ. (Oslo), 1948. 

 Qttart. Jour. Roy. Met. Soc, Jan., 1949. 



Discussion 



In reply to a question b}^ Mr. Andrew Thomson as to whether stations 

 on meridians or on parallels are more useful, Mr. Priestley replied that, 

 ideally, both were required, but in the event of only a few stations being 

 available those on different parallels of latitude would be most useful. 



In answer to a question by Mr. Hutchings, Mr. Priestley said that the 

 unavoidable slight errors in temperature given by radiosonde data 

 were not a serious hmitation in this work. What was required was 

 continuity in time. 



NEW METEOROLOGICAL INSTRUMENTS 



By Andrew Thomson, Controller, Meteorological Service of Canada 



[.4 hstract] 



1. The Meteorological Service of Canada has recently developed the 

 Gill hydrogen generator which will produce about 100 cu. ft. of hydrogen 

 in about eight minutes at a maximum pressure of 3 lbs. per sq. in. 



The process employed is to add water at a controlled rate to a mixture 

 of flake caustic soda and aluminium turnings. The reaction produces' 

 much heat and steam, which is quickly removed by passing the gases 

 through boiler tubes filled with a cooling solution. After the reaction is 

 completed the residue can be washed out easily, and within two minutes 

 the generator can be made ready for generating another charge of 

 hydrogen. 



2. There has also been developed (along the lines of the Robitsch 

 pyrheliometer) a solar radiation instrument which wiU record mechanically 

 the total amount of solar radiation falling on a horizontal surface. A 



127 



