OBSERVATIONS AT KITE STATIONS, 1906-7 15

DISCUSSION OF THE AIETEOROLOGICAL OBSERVATIONS AT THE BRITISH KITE STATIONS, SESSION 1906-1 9Oi.

BY MARGARET WHITE, T. V. PRING, AND J. E. PETAVEL, F.R.S.

[Read November 20, 1907.1

INTRODUCTION.

INVESTIGATIONS of the upper air are carried out a t four British Stations by mcans of kites and balloons.

These stations are- 1. Pyrton Hill, near Watlington, Oxon, 500 ft. above sea-level. 2. Ditcham Park, near Petersfield, Hampshire, 400 ft. 3. Brighton, Sussex, 380 ft. 4. Glossop Moor, Peak District, Derbyshire, 1100 ft.

The work is under the general direction of Mr. W. H. Dines, F.B.S., who has devised most of the special apparatus in use.

The present paper consists of an analysis of the records obtained during the past session.

The kites carry meteorographs which give a continuous record of temperature, wind velocity, humidity, and height. Some 5000 readings hare heen obtained from these records and form the principd hasis of the following conclusions.

Tenyernt w e Giczdients.

Under ordinary circumstances the atmospheric temperatiire decreases The rate of fall is, in the daytime,

The average temperature gradient calculated from the data published

continually with increasing height. very rapid close to the ground and diminishes a t higher levels.

during the present session is given in the following table :-

TABLE ~.- -TESIPERATUI~E GRADIESTS (Degrees Falir. per 100 feet) IX 1'a1:1oss

P I W C Ground-IGJO ft. lG10-32SO ft. 330-.1!VO ft. 4 < ~ b G ; t i O It.

OF EXLAND. Average Values, Session 1906-1907.

Glossop Moor . . . 0 '68 0'50 0:3 -1 0%

Oxshott and Pyrton Hill . 0'43 0'41 0 '36 ...

Brighton . . . . 0.44 0 '30 ... ... Ditcham Park . . . 0.55 0'41 .. .

The decrease of temperature is somewhat more rapid at Glossop Moor and Ditcham Park than in the case of the other two stations. The average temperature gradient for all stations works out a t 0".50 per 100 ft. ( O O . 9 2 C. per 100 metres) between the ground-level and 1640 ft. (500 metres), 0'.43 ( O O . 7 6 C.) between 1640 and 3280 ft. (500 and 1000 m.), and 0'.34 (0"-61 C.) for the next 1640 ft. (500 m.).

In Fig. 1 the average temperature height curves for the various stations are given. For the purpose of comparison curves embodying the

nd L e v e l

FIG. ?.

OBSERVATIONS AT KITE STATIONS, 1906-7 17

results obtained a t Blue Hil1,I Berlin,2 Oxshott, and Crinan during previous years are i n ~ e r t e d . ~

The theoretical fall of temperature due to the adiabatic expansion of dry and also of saturated air mill be found on this diagram.

The Eerlin, Crinan, and Oxshott ciirves n g e e closely with the calculated gradient for saturated air, whereas the Blue Hill and Glossop Moor curves lie closer to the dry air values.

Let us now consider what factors might n pr io r i bc expected to cause a departure from the above-stated average conditions.

Firstly, the temperature gradient varies with the direction of the wind. At Glossop Moor i t is largest for a North-west wind, as will be seen in the polar diagram (Fig. 2).

W

S Flo. 3.

The distribution at Brighton and Ditcham Park is shown in Fig. 3. At Oxshott and Pyrton Hill both the change of gradient with height, and the variation with wind direction, as calculated from this session's observations, are very small.

Numerous observers have noticed that in a general may the tempera- ture gradient is highest on fine cloudless days, and lowest when the sky is overcast.

This led us to calculate the average temperature gradient a t Glossop for a high and for a low barometric pressure, with the following result :-

See Dines, PTOC. Roy. Sue. vol. Ixxvii. p. 443. N e t . Zm't., 1904,,p. 459.

J See also for Pnrls, St. Louis, Blue Hill, Berlin, 011~1 Moscow, Ed,.. z. PIiys. tl.frPie,~ iltjiws., 1906, vol. i i . p. 40.

c

18 OBSERVATIONS AT RITE STATIONS, 1906-7

TABLE II.-AVEMGE TEMPERATURE GIiAnIEST FOn HIGH ,\XD L O W BAROMETRIC PBESSUI:E.. ( D e p c s Falir. per 100 feet).

Gmun,i-iao n. ~ I H O - ~ ~ S O ft. 3?8o-mn n.

hronieter above 30 inrhcs . . O ' s 4 O G l 0% ,, below 30 iiiclies . . 0.59 0'41 0.31

Thc tlifference is small hilt seems to bear out. the statement. If, howevci., we take as our criterion the actual percentage of cloutled sky during the ascent we obtain a much more definite result.

0 v ~ i : i u i s r D i n (Degrees Fnlir. per 100 feet). TAI%LE I~~. -TrS~II 'EI:ATI'I :E GI:AIlIENT AT GLOSSOP 0s CL0UDl.ESS A S D US

Clou~llesa or almoat Sky partly Sky entirely cloudless ~ l i y . cloiided. overczst.

Gr011ud-l64O f i . . . . . 0'91 0484 o h 1640-3230 ft. . . . . 0'60 0'55 0.39 3230.4920 ft. . . . . 0 3 3 0) 0'28 0'27

IVith a cloudless sky, therefore, the average temperature gradient is always large.

This effect is doubtless due to the difference between the adiabatic cqnnsion of dry, sntiiratetl, arid partly s:itiiratecl air ; and this conclusioii is I,orue oiit by a coinprisori Ixtweeii the 1)ercciit;rge of humidity correspoiitling to the aver;rgc of the highest mid lowest temperature grad i e 11 ts.

TABLE: I ~ . - T E l I l ' E R A T T R E Gl:AnlI?NT ASD HUJIIDITT, GLWWP, 1906-7 (Uegees Falir. per 100 feet).

Glw,ic.nt, Corresponding Avenge Humidity.

O/ /" -- Ground-1640 ft. . . . 0 5 9 I 1

0.45 91

1649-3290 ft. . . . .

330-4920 rt. . . . .

0 '63 82 0.36 94

0'46 79 0'17 94

The observations were also analyscd in order to cstim:ite the connec- tion between the wind velocity and the tcrnper:rture gradient. Taking for each height the ;rverage, firstly of thc lowest and then of the highest tenil)er';tture gradients, we obtain the folloming resiilt :-

TAl3LE V.-IYISD VELOCIlT A s n TEMPEi:.ATLXE GKADIESTS. Temiwrature Gradient in degrees Fahr. pcr 100 ft ; Wind Velocity in miles per hour.

I'lnce. Gmiind-1WO ft. 1640-3?SO 3!?SO-.II)M Temp. Wind Temp. Wind Temp. Wind

Gradient. Velocity. Gmdiant. Velocity. Grsdient. Velocity.

G l o s o p Noor . . OP45 17 O h 0 28 0'17 35 0.99 15 0.63 21 0'46 31

Ditclianl Park . . 0'35 30 0'35 30 ... ... 0 .i3 26 0'52 28 ... ...

Briglitoil . . . 0.13 20 0 '09 29 ... ... 0'71 22 0.49 24 ...

OYSllOtt . . . 0.22 22 0'12 23 ... OYi5 I 9 0'54 34 . . . ...

OBSERVATIONS AT KITE STATIONS, 1906-7 I9

There is, it appears, little doubt that n strong mind and a loiv

The results obtained near Paris, Blue Hill, and Berlin, show a consider- this

temperature gradient usually co-exist.

able variation of temperature gradient from summer to winter ; effect exists in England but it is much less marked.

TABLE VI.--AVERAGE TwrmxrmT's CI:,\nIEST IY SUIIIJIRR A S D WINTER (Degrees Fahr. pcr 100 feet).

Croiial-1040 ft. 1G40.330. 2ZS0-4EO. Average for Eiigluiul.

Average for Ecrtiib.

Summer , . . , 0'59 047 0'38 Winter. . . . . 0'16 0'39 0 2 4

Summer . . . . 0.60 0.38 0.36 Winter. . . . . 0'13 0'15 0'16

The result of this relatively slight alteration of the grndicnt with the season of the year is to render the curves of temperature nearly parnllel a t all heights, as, for instance, is shown in the cLqe of Glossop in Fig. 4.

Wind Yelocities.

A serious defect of the investigation by means of kites is that the conditions under which observations can be taken are limited; a calm or, on the other hand, a very stormy day rendering a successful ascent impossible.

To obviate this defect i t is hoped in subsequent years to provide for 1 See Beilr. z. Phys. d. J ~ e i e i z Atmos., 1906, vol. ii. p. 35.

20 OBSERVATIONS AT KITE STATIONS, 1906-7

the use of balloons on calm days, while the recent installation at Glossop Moor of new machinery, purchased with funds awarded by the Govern- ment Grant Committee of the Royal Society, mill, i t is Iioped, considerably increase the chances of a satisfwtorp 'ascent in very high winds.

I n T;rble YII. the average mind velocities a t various heights have I)CCII worked out from the results o k i i n e d at each of the kite shtions.

'l"\l;r,L \ ' ~ I . - A ~ K I : A G I : \ \ -ISD VEI.OCITII:S (Miles pr I IOIV) . PI3cr. GroiiiItI. IIJII) ft . XYO pt. WII rt.

ClOSSOII altJlII' . . . . 13.01 "6'1 30 '2 33.3

Erigliroii . . . 113.1 ' 213'6 "5 .3 ... Oxsht,tt arill llyrtoii H i l i 13.0 30 0 31 'Y 36.T

1)itcIiaiiil'ark . . . . 24 .T 29 .s 31 2 3i.5 ( I )

1 E4l!ll:ltel.

The average wind velocity at eacli height is tlierefore nearly the samc at nll stations. I t has been pointccl out t h a t these nrimbers do not iiicliitle nll possii)le wxrther corditions ; h i t :is the m;ixiniii ; t ~ i c l minimcr iviricl velocities Iiappcn to he both excluded by esperimentnl difficulties, the error is to a certnin cstent compensntctl. The r:tte of increase is, however, p rohbly soincwhat exiggerntcd, owing to the fiict that the liighcr nltitndcs itre nttaiiiccl only when the upper wind is of considernl,le

l ' hc nveragc wind velocitx increases with height very rapidly for the tirst 8 ; : ~ ft. and more slon.ly a t liiglier levels.

A t the Rlric Hill Olmrvntory the increase in velocity with height is 0.6 mile per hour in slimmer: :lilt1 2.1 niilcs pcr hour in winter for each 830 ft. rise.

111 l.:ngland, near to the groiuid level, a similar change occurs, the wiiicls i n winter increasing twice :is mpitlly as in summer. A t heights :tl,ove 16 4 0 ft. the grntlient still alters with the season. T h e variation, however, is in the rcvcrse ilirectiori, the increase ill winter being 0.3 mile per hour per 330 ft., falling in the summer months to 0.6 mile per hour.

tha t tlic wind vclocitp a t any :rltitride is inversely proportionid t o the density of the atmosphere at tha t altitiide ; that is to say, the mass of :Lir passing :I given point ill a given t h e is constniit for all altitudes, or, expressed nlgelraic:dly,

Vtl= constant

where V is the mind velocity and tE the density of the atmosphere. The law applies wcll for high IcvcIs, b u t at the moderntc hcights

reached by kites thc incrense is ropiJ, and follows more closely the inverse square of the density.

TVe now come t o the question of the alteration of wind direction with height. As is well known from the study of the motion of clouds, and from the data obtained by means of balloons, the direction of the wind d t e r s a t high levels, rotating in a clockwise direction ; for instance, a Soiitli wind tends to become more II'esterly.

I n indivitlual experiments the amount, and even the dircction of the rot:ition, differ widely. Taking, however, averages from the GOO

s t rltng t l l .

I t hns 1)ccii suggcsted by Clayton ant1 by Egnell

1 C ' h p k ~ R c ~ J I ~ I ( . s , 1903, vol. crsxvi. pp. 355-361.

OBSERVATIONS AT KITE STATIONS, 1906-7

observations of wind direction, which are available for the preserlt session (balloon asccnts included), we obtain a consistent result, sliowiilg that up t o 8300 ft. the angle between the wind a t blic ground level and the upper wind is nearly proportional to the height.

21

Height. Tot31 anfile of robtion.

Ground t o 1640 It. 11" ,, 3280 ,, 18" ,, 4920 ,, PO"

,, 8500 ,, 40"

Prcvioua investigatioiis have shown that both under cyclonic n~lt l mticyclonic condi tiorrs tlie change of direction is grcntest in the cnsc of ;L Sonthcrly wind. Talciiig from our observations tlie average rotation for siich winds IVC obtain angles of 13", 35", and 67" between the wind ;It the groiind level, and a t 1640, 4980, and 8200 ft. respectively, ic., a rotation nearly 50 per cent. above the general average.

The invcstigution of tho relation between mind velocity and barometric p d i e n t may not seem a t first sight to offer serious difficulties. 111

,, 6600 I , 30"

FIG. 5.

Fig. 5 the gradients and resulting mind velocities, as measured a t Glossop diiring Augnst 1906, are plotted, and the synchronous rise and fall of the curves is a t once apparent. It is equally evident that the mind velocity cannot be expressed as a simple function of the gradient. Theoretical considerations confirm this view. The wind velocity depends, not merely on the absolute value of the gradient, bilt on the angle which the wind forms with the direction of the gradient, on the curvature of the pnth, and doubtless also on the wind direction, season of the year, n m l other factors.

A preliminary study of the question showed that the data available were too incomplete to form the basis of reliable conclusions.

22 O B S E R V A T I O N S AT K I T E STATIONS, 1906-7

The exact gradient a t the time a t which the observations are taken cannot be obtained with accuracy ; the general disposition of the isobars over the country is recorded morning and evening by the RIeteorological Office, but i t is only by interpolation that the gradient a t any other time of day can be determined. Again, small local variations in barometric pressure cannot be obtained from the weather charts, but have probably an important effect on the prevailing wind.

It mould be useless to give a further account of this incomplete and therefore unsatisfactory part of our work.

During the corning year i t is hoped to arrange for barometric observations to be made, during the kite ascents, at various places in the neighhourhood of the Glossop Station, with a view to the further study of this question. -

A P P E N D I X The values given in Tables 1.-VII. are reproduced in the Centigacle and

Metric scales in the following Tables for the use of Continental observers :-

T A B L E ~.-TEMPERATURE GRADIENTS (Degrces Centigrade per 100 metres) I N

VARIIWS PAI:TS OF ENCLASLJ. Average Values for the Session 1906-1907.

Plncc Ground level- 500-1000 1000.1500 1500-2000 500 metres. metrrs. metres. metres.

Glossop Moor . . . 144 041 041 0.>3 (?) lhighton . . . . 0'81 0'5i ... ... Ditcham Park . 1.05 O T 5 ... ... Oxsliott and Pyrtoii Hiil 0. i8 0 . i 5 ... ...

TABLE II.-AvER.~GE TEMPERATCRE G I L A D I E S T F01: Hicu AXD Low BAROMETRIC PILESWRE (Degees Centigrade for 100 metres).

Groiind. ~ x ) - l o o O 1000-1500 9 3 metre% metres. metres.

Ihrometer above 762 mms. . . 1'1 i oP-- 1 3 o k I3orometer below i6? mms. . . 1'0Y O ' i 5 0.56

; T i I I ; L E III.-TEJIPECATL'RE GRADIEST AT Gl.osS~,l' n s CI.OUPLEPS A N D ON OYERCAST DAYS ( D e p e s Ceiitigratlc Iicr 100 iiietrcs).

Cloudless or d i i i o ~ t Sky partly Sky entirely cloudless sky. rloudl.ll. overcnst.

Gmnnd-500 metres . . 1 's6 1 .>3 0 .%3 500-1000 nictres . . 1.09 1 .oo 0.72 1000-1500 nietrcs . . 0.60 (?) 0.51 0.49

TABLE I V . - T E M I ~ E C A T ~ I I E CI~ALJIENT ASD ~ I u ~ i l n i ~ r ~ , C; t.ossor, 1906-7 (Degrees Centi@c 1m 100 metres).

Corrrs1nncliny Avernjie Humidity. Gradient.

x ~round.500 metres . . . I'SO 77

0.82 91 600-1000 metres . . . 1.1-1

0.65 1000-1500 nietres: . . . 0.83

0'31

82 94 79 94

OBSERVATIONS AT KITE STATIONS, 1906-7 23

TABLE v.-\vIXD VELOCITY AND TEMPEnhTtinE GRADIE?ITS.

Temperature Gradient in degrees Centigrade per 100 metres; Wiud Velocity in metres per second.

500-1000 metres. 1000-1500 metres. Grounil-500 metres. Temp. Wind Trrnp Riiid Teitip. Wind

Place. Gradient. Velocity. Gmdiunt. Vulocity. Gradient. Velocity.

111. p. 9. 111.p.Y. m.p.s. Glossop . . 0082 i .6 0.%6 12'5 0 . h 15'6

1.505 6.7 1'13 10.7 0'34 13'8

Ditchani Park . 0.695 13.3 0.645 13.1 ... ... 1.326 11.6 0.943 12.5 ... ...

Brighton . . 0,232 8'9 0'168 13.1 ... ... 1 2 9 7 9'6 0.890 10.7 ... ...

Oxshott . . 0'391 9.8 0.186 12'3 ... ... 1~183 8.3 0'980 15 ... . . .

'rA u L E VI. -11 YE 1:AO IS TEMPE I: AT u 1:E ci ILA D I ENT I 9 SU >I JI E I: AS D li'l STll R

(Degrees Centigrade per 100 metres). Ground- y0-1000 lMX)-ISM)

500 metres.1 .iiietrcn inetri:s. Average for England.

Snmmer . . . 1'08 0.56 0.70 Winter . . . . 0'84 0 .ti9 0'43

,1 ccrage f b r Bcdin. Siimincr . . . 1.09 0.69 0.65 Winter . . . . 0 2.1 0.27 0'30

TABLE VII.--~VERAGE Wrsn Vmt)ci.ri!?s (J[etres per sccond).

Glossop Moor . . 5.8 I 11'6 13.5 14.9 Oxshott and Pyrton Hill 5.8 13.3 14.2 16.4 I3rigliton . . . 7.2 1 11'8 12.7

Place. Grontid. 500 inetres. loo0 metres. 1500 metres.

Ditchan1 Park . . 11.0 13'1 13'9 16.7 (?)

1 Estimated.

DISCUSSION.

Jlr. W. L I A ~ ~ I O T T said that the Kite Committee had considered it desirable that balloon observations should be taken in Ireland. Sir John Moore hat1 suggested that they shonltl apply to Prof. W. E. Thrift of Trinity College, Dublin ; and he, with the assistance of the Astronomer-Royal for Ireland and other gentleiiieii, had entered into the matter very heartily. The Kite Com- mittee considered it not desirable to send up any nieteorographe, as Dublin was so close to the 663. A number of pilot balloons were used, and arrangements were made for having two or three observers with theodolites, to get the altitude and rate of travel of the balloons. The great difficulty experienced was with the weather. The sky was covered with low cloud, in which the balloons were lost in less than a minute. Only one of the balloons hail passed over the sea, and that fell in Wales. JIost of the balloons fell near Dublin, and did not attain any great altitude. A great deal of interest had been aroused, and much atten- tion given to the subject in Ireland.

"4 DISCUSSION-OBSERVATIONS AT KITE STATIONS, 1906-7

Mr. C. J. P. CAVE said that he Itad sent up six hlloons in July, of which lour had been returned. He showed the temperature and height climes on a lantern slide. Some of the balloons sent up in the daytime showed an undue rise of tcinperatare a t the highest points reached. The balloon sent up on July 23 gcive higher temperatures than usual, due perhaps to solar radiation, as the balloon WNI probably not ascending fast ; i t was leaking and mas found in- tact next day floating in the Channel. A balloon sent up on the evening of July 2-1 went up to 52,500 feet (16,000 metres), and mas matched with one theodolite for two hours. The sun set on the balloon 22 minutes later than it set on the ground. July 25 was the only day on which an ascent was made with an iiistruruent other than that of Jlr. Din-', one of U. Teisserenc de Bort's meteorographs being used. The balloon of July 27 was the only one sent u p at iiiglit ; it reached a height of 23,000 ft. (8500 metres), and the tlecrcase of tcniperature w.u rather greater than iii tlie otlter cases. He was very grateful to JIr. Dines for the instrument he Iind designed, as those used on the Continent were lteavier and much niore espensivc. I I I the International ascents En:lnnil headed the list for height with two .scents a t Manchester of ti!),OOO feet (21,000 metres), and out of nine asceiita of over 65,000 feet (20,000 ttictrcs) four took plncc in England.

Captaiti C. 11. LEY mid that his observations h a d becn cliiefly carried 011

with pilot b;illoons, of which lie had sent up tiiore than thirty. He ltad nsed a iitctliod tlilFerciit fIotii thc otlter itietliods, consistin; of niensuring tlie diameter ol' t.lie ballooii I J ~ wires in tlic diapltr:ipi of a tlieodolite, and froni t1t;it getting tlie range of the balloon. Observations of tlic altitnde deteriiiiiied tlie vertical lieiglit and tlic horizontal distance. Most of the results had beeti plutted, and lie w.19 subniitting a general report on tlteni. They were interesting chiefly in tlie matter of vclocitics. Tlicy showed no steady velocity, and the average over a 1011; run coiilJ give only fair results, but there were considerable and sharp variations. His results agreed generally with the few retorns given by Jlr. Dines' instraincuts of the tenipeinture variations. When the yelocity wu f a t the teiiipcraturc gradient was slom.

Ur. W. N. S t u w reniarked that it w . 1 ~ not long ago since the investigation o f the upyier tiir ltad becn put aside, on tlie ground that it was it~ipossible in t h i s country, atid it w3s extreniely gratifying to find, within n few years, that the upper iiir lind been investigate(1 by the scent of about twenty-five Idloons dwing a single week. -4 new system of observations had also becii introduced Ly the niethod of pilot balloons and their examination by theodolite?. Caytnin Ley liacl used this nietliotl with reniarkable effect, and it promised to form tlte subject uf an interesting paper a t a Inter date. They had Iieard two s e h of papers, the first conticctcd with the international ascents in tlie last week of July, a1111 the secottd coiicerning the gctieral results of kite and balloon observa- tions, contparing the results obtained a t different etntioris i n the British Isles during tlie year 1 3 O G - i . I t was rarely that any one was found to discues the obsermtioiis ttiade by other people, and they Itad to thank Mr. Petavel for so doing. JIr. Petavel had dealt with the results of tlie differelit kite stations by the nictliod of iiieati values. Wlien cotisidering the observations returned to the Jlctcorological Ofice with blr. Gold, he lid thought i t desirable to go bcltind this n1ctlioc1, aud to etideavour to make Eome pictorial representation of tlie resiilts of the kite observationp, in which the classification of tlie observations accordin; to frequency W ; L ~ represented. The diagrams lie had ltad prepared were included in a paper r e d before the Aeronautical Society 1 s t December. He showed four slides, which gave the results a t Oxshott in the form of differences at successive steps of 500 nietres (1640 feet) froni the values at 500 nietres above sen-level. Dr. Shaw theii called attention to the confirmation which the recent csperimeuts alforded to the existence of the stratosphere, a layer of air

DISCUSSION-OBSERVATION9 AT KITE STATIONS, 1906-7 25

above which there would only be a slight variation of temperature. This isothermal layer waa found at about 10 kilometres (6.2 miles), and its existence wag clearly proved by many ascents. The curves compared satisfactorily with the best of those obtained on the Continent. aS to the origin of this layer there were different suggestions, among them the question as to whether the propor- tion of gases did not vary. M. Teisserenc de Bort had endeavoured to mertain the facts by using vacuum tubes, which were opened and sealed automatically at great heights.

Mr. J. E. PETAVEL remarked that the high temperature gradients obtained at Glossop Moor over the first height interval were doubtless due to the altitnde of this station, which is 1100 feet (345 metres) above the sea. The first gradient is therefore the difference of temperature per 328 feet (100 metres) over a height of 492 feet (155 metres) only as compared with an average interval of 1310 feet (400 metres) for other British stations. As is well known the gradieiit very near the ground level is always exceptionally great.

The PREYIDEXT, in conclusion, &I they woulcl all unite in conveying their thanks to the authors of the reports, Jlr. Dines, Jlr. Petavel, Jlr. Harwood, Prof. Thrift, and Miss White, through the gentlemen who had read the papers. Captain Ley's report would, he hoped, be heard on a future occasion. Time did not ndniit of further speaking, but he niight add that the nienn ciirves of the vertical temperature gradient in the air were reniiiiiscent of the typical curves of the vertical teinperature gradient in the ocean and in sene, the region of rapid change near the surface passing gradually into an isothermal (or as he had ventured to term it, a honiotheriiiic) layer. I n terminating the inceting he announceil that an interesting Model of a Cliiuatological Station, prepared by Jlessrs. Segretti and Zaiiibm, was being exhibited in the ten-room.

J a m w y 16, lDOS.--The :iverqc height of tliese ascents is over 13,000 nlatres, :lnd n region where tlie temperature change with height is trilling lias Iweu rcwhed in 26 c;wes.

Hail Shooting in Italy. In tlie United States .\[o?~t/dy PVenther It'euiew fur -ingust 1907, there is R

Note by Professor Cleveland Abbe, in which lie says that tlie references to this subject in previous volumes of that publication have abundantly shown the prob- ability that there is no rational basis for the efforts made in Italy and France to break up thunderstorms and prevent injurious hail by some method of cannonading. Neither the noise, nor the smoke, nor the heat, nor the coin- motion produced by grand vortex rings om be expected to have any consider- able influence on the enornious cumulus clouds from which hail and lightning proceed. This conviction is now confirmed by a report read before the Royal Academy of Scicnces at Rome (Accademin dei Lincei), on Deceiiiber 2, 1906, by Senator P. Blaserna, who is also Professor of Physics in the Royal University at Rome, and President of the Accadeinia dci Lincei. In 1902 Professor Bhserna was appointed by the Italian Government president o l R special com- mission to investigate the subject. A locality that had suffered extremely in previous years was clioseil as the field of operations, viz. Castel-Franco in Venetia, and 22% cannnn of tlie most approved special type manufactured by the Greinitz Company were established ; each of these sends up a vortex ring 13 feet in diameter, and one additional cannon sending up a vortex 45 feet in diameter was subsequently added. As these vortices failed to ascend higher than 200 or 300 yards they evidently had no effect on the clouds ; therefore a higher station, the Casa Aulagne di Monteux, was occupied, so that the vortex rings attained 1200 yards, but still no good results were perceived.

26 HAIL SHOOTING IN ITALY

Then the Secretary for War and the manufacturers of pyrotechnics were appealed t.0. Of the latter, Marazzi, a t Rome, succeeded in constructing bombs weighing 18 lbs. that were wried up to 2500 feet, where they exploded. During 1906,250 broadsides were fired by the 222 w o n at Aulagne, and 60 of the 3h'aZZi bombs were sent up, but still no good effects were perceptible. These negative results of a five years' campaign justify the commission in re- commending that the Italiau Government no loiiger encourage such expensive and useless work.

Audibility of Clock Bells. In connection with blr. Jlarriott's notes on sound (and the clock bell of

St. Paul's Cathedral being heard a t a distance), in the Q~urr ter ly J o w n l of the Royul dfeteorological Society, vol. XI. p. 243, the following extract from iYotes a id Qz~enks (November 16, 19Oi) is of interest :-

The great clock bell of St. Paul's possesses the peculiarity of sounding tligerent notes from different positions and distances (of different degrecs of riudibility). The architect who superintended its recasting must have forgotten to provide a section, consequently the bell is " straight-waisted," which accounts for its toleestain sound. All curves in the section of a bell should be sections of a cone, and of the same cone. Probably (says Vr. Walter Scargill) the architect niay have forgotten to read his Vitruvius, who shows that an architect must not be " ruusicdeaf."

Influence of Temperature on the Iron Bailway Viaduct at Crumlin. Begun in 1853, the g m t viaduct in ~Io~imouthshire wcw opened in11857, but

after its completion temporary dificulties were experienced owing to the expan- sion and contraction of 90 vast a body of ironwork, and the girders had to be supported on rocking links.

The records which were taken show a variation, owing to the difference of teuiperntore, of of an inch in the length of the viaduct between 4.0 a.m. and 4.0 p.m. on June 12, 1861, n line warm day following a cold damp night. The greatest difference in length agte;;ates over two and a half inches, the extremes occurrins respectively on February 12, 1861, when the temperature was 32", and on August 27, 1861, when the teniperature was OO", a difference of 58".--Great JVestrriL Ruilzcccy Nnpn'ne, October 1907.

The Furness Railway Wind-Gauge. More than two years ago aome allusion ivaa made (Qitnrlerly Journal of

the Ruynl Xeteoroloyical So&@/, rol. xxxi. p. 189) to the wind-gauge erected in the neiglibourhood of Ulverston to protect trains from risk in traversing a very exposed viaduct. A full description of this is now given in The Bailway J1qcc:ine (October 1907, p. 337, ct seq.) with an illustration. The apparatus is tixed at the west end of the Levens viaduct, and its actuating part consists of two buardq kept in a vertical position by springs, and the movements of which are shown on a chart by means of the customary pen and clockwork appliances. A pen is operated by either board according to the direction of the wind, and for greater exactitude of time, the chart paper, over sixty-five feet long, is perforated, the holes in the paper corresponding with studs in a clock- work wheel.

When the wind pressure reaches 32 lbs. to the square foot the spring boards idcrrcd to are adjusted to make an electrical contact, r i n , ~ g bells i n the signal cabins on either side of the viaduct, and when this occurs all trains are cletainecl until the force of the wind motlerntes-any such interruption beiug tele,ornphetl to the superintendent of the line. The apparatus is tested weekly. It has I ~ e n at work since September 1903, and in February 1907 a velocity calculated as etliiivalent to 65 uiles an hour was recorded.