864 



NATURE 



[Decembi 



" mixing layer," for it contained a much smaller 



iwoportion of water vapour per mas* of air than the 

 ayers above and below. The following considera- 

 tions suggest that it belonged purely to the lower 

 current and was, indetxl. tiie only genuine " polar 

 air " surviving over CranweU. 



Let the shaded figure (a), in the inset diagram, 

 represent, schematically only, a vertical section 

 (vertical scale much magnified) of a " tongue " or 

 " u lobule " of cold air which has not long since 

 l>( iictratcd from a cold source into a warmer environ- 

 ini lit. Subsequent translation and lateral spreading 

 under gravity may lead to a condition, now at a 

 distance from the cold source, represented by fig. (6), 

 the line-squall taking place, say, at the right-hand 

 edge. A contribution to the study of such motion 

 has been made by Exner [Sitzungsber. Akad. Wiss. 

 Wien, Ila, 127, 1918, pp. 795-847I. In the transition 

 the air at A descends to C. while the surface air 

 remains at the surface, so that the higher a layer was 

 originally the more it is warmed adiabatically. Hence 



Fig. I. 



if the original lapse -rate, represented by AB in 

 fig. (c), is less than the adiabatic, as is reasonable 

 for air cooled mainly from the surface in high latitudes, 

 it will become, from this cause alone, more remote 

 still from the adiabatic, like CB, where AC is the 

 slope of the " dry adiabatic." A more precise 

 mathematical treatment of such a change of lapse- 

 rate has been given by Margules (Exner, " Dynamische 

 Meteorologie," 1917, p. 80). Any change of pressure, 

 at a given level, during the transition would produce 

 effects which would not appreciably modify that 

 demonstrated. The sinking of the globule thus tends 

 towards greater stability and therefore to oppose 

 turbulence. But in passing over a relatively warm 

 surface such as the Atlantic the cold air mass would 

 receive heat from below, and this would be diffused 



NO. 2824, VOL. 112] 



upwards as far as the turbulence originating at ti.t 

 surface is able to penetrate in the face of the stabiL^ 

 ing action of the sinking. A lapse line like CI I) 

 would result, where E b tne upper limit of mixing' 



In the present case there is clear evidence of th.- 

 upper limit of turbulence in the " haze-top" report' 1 

 at about 4000 ft., and the layer ED in the m.iii 

 diagram was accordingly the one which had b<-< n 

 warmed from below ; it possessed the " dry adiabatK ' 

 lapse-rate of temperature indicative of thorough 

 mixing, and was fairly humid, with cloud at the top, 

 owing to water vapour, evaporated over the Atlantic, 

 having been stirred up at tne same time as the layer 

 was warmed. The layer CE, on the other hand, was, 

 on this view, one which had succeeded in r<' 

 non-turbulent, and was really the only 

 " polar air " which reached CranweU, still posscssnig 

 the original low polar water vapour content and. 

 neglecting radiation, its original potential ♦,-——* 

 This view is strengthened by remarking t : 

 the moderate decrease of humidity and :^iii,iii ii>t; w. 

 temperature indicated by the upper parts of the two 

 records may be explained, using a Hertz diagram, 

 by supposing the " equatorial air " to have ascended, 

 lost some moisture as " cold front " rain, and de- 

 scended again, as suggested at F in fig, (6), no such 

 explanation can be applied to the extremely dry layer 

 discussed above, for the air would have to be taken 

 to an unreasonably great height, and would also 

 arrive back with far too high a temperature. It does 

 not seem, therefore, that the intermediate layer can 

 have been evolved out of the " equatorial air," and 

 these interesting records are accordingly to be inter- 

 preted as showing that the air of .p>olar origin over 

 CranweU was in all probability a " tongue " or 

 " globule " about 9000 feet thick with the lower half 

 partially " dei>olarised." Subsequent weather charts 

 suggest that it was soon replaced again by a warmer 

 air current right down to the surface. 



Such pairs of records as this, near " cold fronts," 

 are uncommon, the nearest approach amongst those 

 published in the DaUy Weather Reports being in the 

 issues of October 14 and 15, 1921, two ascents at 

 Baldonnell (Dublin) having exhibited simUar features 

 to the present. They are also important as bringing 

 into prominence the existence well within areas of 

 cyclonic activity of " dry inversions," the occurrence 

 of which is more commonly associated with the 

 margins of anticyclones. M. A. Giblett. 



Meteorological Office, 



Air Ministry-, Kingsway, W.C.2, 

 November 14. 



Experiments on Ciona intestinalis. 



It is remarkable that in all the statements that I 

 have seen by Dr. Kammerer or Prof. MacBride con- 

 cerning the increase in length of the siphons in Ciona 

 following amputation, no measurements are given. 

 As I have recently made measurements on a number 

 of specimens in order to obtain some data indicating 

 the natural variation in the proportional length of 

 the oral siphon, I should be glad if Prof. MacBride 

 would inform us whether Dr. Kammerer has published, 

 or whether he possesses unpublished, any measure- 

 ments of the siphons in his specimens before and 

 after amputation, and in the offspring which in- 

 herited the increased length. 



I find it difficult to understand what Prof. MacBride 

 means by the words " the reaction is of the animal 

 as a whole." It may be a fact that amputation of 

 both siphons results in the growth of longer siphons, 

 and of the oral siphon alone in the growth of a siphon 

 of the same length as before, but I do not see how 



