608 TBANSACTIONS- OP SECTION A i 



iu this respedt uutil the various conditions which actually prevail iu the atmo- 

 sphere have been more thoroughly explored. Fortunately, the proj^ress which i» 

 now being made in the study of the upper regions of the air encourages us to 

 hope that before long we may have more secure data to form the basis and the 

 criterion of future theoretical investigations. 



Dr. Shaw expressed his appreciation of the clearness of Professor Lamb's 

 paper and of the advantages for the study of the atmosphere likely to accrue from 

 having the results obtained by mathematical investigation put together in a 

 concise and easily accessible form. So far as the atmospheric oscillations of short 

 period were concerned it was not easy to accept an explanation depending upon 

 the assumption of a substantial discontinuity of velocity between consecutive 

 horizontal layers. There was plenty of evidence of stratification in the atmo- 

 sphere of one sort or other, but pressure, being an integral gravitational effect, 

 could not show discontinuous stratification. Consequently there could not be 

 a discontinuous change of horizontal pressure-gradient. There might be slight 

 discontinuous stratification as regards temperature, and consequently also as regards 

 density ; but, as regards wind velocity, anything of the nature of a persistent wind 

 current in tbe upper air would have its velocity governed by the pressure gradient 

 and the density, and the discontinuity would be limited to the extent of that 

 shown in the density, which can very rarely reach so great a proportional 

 magnitude as 3 per cent. 



The records obtained from autographic instruments occasionally showed 

 undulations, more or less sinusoidal in charncter, which had not hitherto been 

 explained. Neither the proximate cause of the initial disturbance, nor the 

 oscillating system which was set in vibration thereby, had bean identified. Such 

 approximately sinusoidal oscillations were sometimes found in the wind, as shown 

 in the example taken from the South port Observatory records for January 12, 1907 

 (see Plate VII.), but more frequently in the barometric pressure, and particularly 

 in the records of the ' inicrobarograph,' an instrument which, as ordinarily 

 adjusted, magnifies rapid variations of pressure twentyfold, as compared with 

 the mercury barometer, but disregards slow changes altogether. A period of about 

 twenty minutes is frequently to be identified in the records, and the trace fron) 

 Mr. Dines's microbarograph at Oxshott for February 22-23, 1904, reproduced in 

 the original paper,' shows intermittent groups of oscillations lasting for about 

 twenty hours, commencing with a period of about half an hour and ending with 

 a period of about ten minutes. 



Barometric oscillations were occasionally set up during the violent commotion 

 of the line squall which introduces a thunderstorm. A good example is given 

 in the microbarlc trace for August 2, 1906 (see Plate VII.), where an oscillation 

 with rapid damping is clearly indicated. The changes in the other elements are 

 reproduced in order to indicate the general violence of tbe commotion. 



But ' line-squall ' or thunderstorm changes do not always cause oscillations. 

 The microbarographic trace for June I, 1908 (see Plate VII.), shows a sudden rise 

 of pressure as usual (the crochet d'orage or Gewitter-nase), but no recognisable 

 undulations ; although the sudden violence of the wind is very conspicuous in the 

 records, and was still more so in reality on account of the destruction of trees caused 

 by it in Bushy avenue. 



Another example was given of violent barometric fluctuations on July 12, 

 1908, wbioh were recorded during thundery weather, but without producing undu- 

 lations or indeed any other notable effects. 



On the other hand, there were examples of regular undulations set up without 

 any considerable barometric change, the most noteworthy of which was the succes- 

 sion of four undulations, commencing with a range of about five-thousandths of an 

 inch, lasting about a quarter of an hour, and then violently interrupted by a sudden, 

 though slight, explosive disturbance which set up different and much faster oscilla- 

 tions for a similar interval. Traces of this train of atmospheric waves from five 



' (p.J. Roy. Met. Soc, vol. xxxi. 1905, p. 39. For a discussion of the theory 

 of the instrument, see Chrystal, Proc. H.S.E., vol. sxviii. 1908, p. 437, 



