376 



KNOWLEDGE. 



October, 1911. 



temperature. The scale is such that one-tenth 

 inch equals lOOC. sd that TC. or -001 inch can 

 easily be read with a low power microscope. A 

 thin strip of German silver is used with the rod 



^Investi^aUon cf the Upper ft ir /SOTS 



D-rj^ar^ Park A 

 Til! c'rx't^ '■OKjr-c/ rt-t sarcrtr/ afa^iofK '^'tACf** a 



Corr-tiponcl.ng dafti art •nrf>coft^ bj '*'l iarr-m 



€^ 



As to accuracy Dines is correct for temperature 

 within 0-8-C against IT by the Continental records, 

 hut the height records are less reliable, the a\erage 

 pressure error being eight millimetres or one-fifth 



inch, against fi\-e millimetres, or 



rather under (Hie-sixth inch. 

 Although at three miles these 

 errors are only about one hundred 

 and fift}' yards, at nine miles the\- 

 increase res[)ccti\el\- to about five 

 hundred and fifty and four hundred 

 and thirt\' yards, and at twelve 

 miles to over one thousand one 

 hundred (one kilometre) and six 

 hundred and seventy yards. 



Finally we turn to the object of 

 all this work and the measure of 

 results attained. .\s to the former 

 there is now added, so far as the 

 first two miles or so is concerned, 

 the additional desire of knowledge 

 helpful to aviation. This has 

 alread}' borne fruit ; for it has 

 been established that eddies and 



uprush gusts.- 



all dangers bv the fliers, - 



By the iojirt^iy o/ 



FiGURK 5. 



Uis Mnjcity's Statioficrv Ojffice. 



Map of British Stations. 



of invar, the latter being a non-expanding allo\-. 

 The two are slightly separated, and fixed at one 

 end. The pen lever, attached to the free ends, 

 multiplies the difference of expansion some tenfold. ' 

 The whole is encased in an aluminium c\linder, 

 open at each e;id. 



Because of this extreme lightness, English results 

 include a larger proportion of records up to tweKe 



miles or so, which is some set off, both to the these Upper Air explorations, to which reference 

 omission of humidit\- records and the greater danger mav now be made, 

 of loss by falling into the sea. It has been, of course, long known that 



-most dreaded of 

 -rapidly 

 decrease as the surface is left. 



But the original incentives, 

 other than purel}' scientific, were 

 more entirely meteorological and 

 practical, natnely. to enable the 

 authorities responsible for weather 

 forecasts to attain a yet greater 

 accuracx'. So long as surface ob- 

 servations only were available, any 

 complete discussion of the causes 

 £5^ and modes of air movements was 

 /^5"\s plainly as iinpossible as it would 

 ^^ be for a geographer thoroughh" to 

 survey a country without lea\ing 

 a railway running due north and 

 south. I'or just as, to him, east 

 and west observations are equally 

 as essential as north and south 

 for his two-dimension chart, so 

 also does the meteorologist require 

 to work in three directions for 

 his three-dimension chart. Already 

 this has been strikingly illus- 

 trated bv Dr. Shaw, the capable head of our 

 Meteorological Office, who built a most beauti- 

 ful and instructive model, in the form of a 

 triangular prism, of the atmosphere to a height of 

 fifteen miles above the British Isles, the angles l>ing 

 at Limerick, Crinan and Pyrton Hill in Oxfordshire 

 (see Figures 1 to 4). A casual glance reveals the chief 

 and least expected phenomenon brought to light by 



This description follows the construction of the hghter apparatus, as used in 1910. 



