Project Cirrus Progress Report No. 5

8/9/2019 Project Cirrus Progress Report No. 5

http://slidepdf.com/reader/full/project-cirrus-progress-report-no-5 1/23

http://www.blackvault.com/



UNCLASSIFIED I LIMITED ADB967

Project Cirrus. Occasional Report Number 5

GENERAL ELECTRIC CO SCHENECTADY NY

15 SEP 1948

Distribution authorized to U.S. Gov't. agencies and their

contractors; Administrative/Operational Use; 15 SEP 1948. Other

requests shall be referred to Signal Corps Engineering Lab.,

Belmar, NJ.



UNCLASSIFIED I LIMITED

Redistribution Of OTIC-Supplied Information Notice

Al l information received from DTIC, not clearly marked "for public release" may

be used only to bid on or to perform work under a U.S. Government contract or

grant for purposes specifically authorized by the U.S. Government agency that

is sponsoring access OR by U.S. Government employees in the performance of

their duties.

Information no t clearly marked "for public release" may not be distributed on the

public/open Internet in any form, published for prQIfit or offered fo r sale in any

manner.

Non-compliance could result in termination of acc'ess.

Reproduction Quality Notice

DTIC's Technical Reports collection spans documents from 1900 to the present.

We employ 100 percent quality control at each stage of the scanning and reproduction process to ensure that our document reproduction is as true to the

original as current scanning and reproduction teclhnology allows.

However, occasionally the original quality does nc)t allow a better copy.

If you are dissatisfied with the reproduction quality of any document that we

provide, please free to contact our Directorate of User Services at

(703) 767-9066/9068 or DSN 427-9066/9068 fo r refUind or replacement.

Do Not Return This Document To OTIC



i==

1<

l1lQD-B967 348

-

O C C ~ J i P . . ! l I . U J ~ . L j J ..- PROJECT CIRRUS

, C O . ~ ) t 9 . ! . L ~ L - W . : ~ ( ' - -I

DTlCELECTE D"

II S OCT '

I,

' - . ._ ..... ._--_ .... ..- _.. . . . . . .. . . .

SCHENEC1ADY, NEW YORK

l5 September :1948

_n__ _ _ _ 0_



--- ---.-------,c - - : : - - -- - -- - --- - '---

SIGNAL COllPS , ENGINEERING L"ABOltATORIES

EVANS SIGNAL. LABORATORY

,.1ETEOROLOGICAL BRANCH

BELt,\AR, NEV, JERSEY

Distribution List for :Reports

fromG e n e r ~ l Electric Research Laboratory

onPROJECT CIHhU&

(Contract No. W.36-039-6C-32427)

No. of

Copies Organize t ion

2 Chief, Research SectIon, Meteorological Branch, ESL 1 Chief, i.leteorologic:al Branch, ESL

,1 i))eteorological BraJ1ch, E ~ ~ L (contra.ct file) 1 Bil eeter of E h g i f i C E ~ I ' : ' l l g , 3GEL

1 Chief, Research Bl'f:lnch, Headquarters, SCEL , 1 T e c h n i c a ~ Repcl'ts Library, ESt.

3 AGF' Lil;lison, r.,sL

2 USAF' Liaison, \ ~ e a t h e r , SeEL

j,lail & Records, ESL

10 Chief, -E!ngineering_&. Technic:al D1V., SlGGE-h1-1, OCSigO, Vashington. 25, D.-C.

-2 Chief, Enj£ineering & Technic£!,.l DiV. ,1 SIGTI:l-S, OGSigO, WashinGton 25 , D. C. I Chief, Research &: t:ngineel'in5 Division, Office Chief' Chemical Corps,

Army C h ~ m i c a l Center, l',ld. ' Att: D:['. 'n.. L. Fox

1 Office pi ' the Quartel"rntl.ster tjeneral, \ 'ar Depurtment , Washington- 25 , D. ') ./1 Office of C h i e ~ ' of O r ~ i n a n ( ' . e , riashingtcn, D. C., Att: Or. 1'. H. Carten,

ORU'fS, Room 1.0436, Pen ta.gon Building

1 Director, Pitman-Dunn l"aboratol-Y, !'ra.nki'ord Arsenal, Philadelphia. 37) Pa.Att: r:lr. J. w. }.\itchell

1 Hollman ltir Force Base, Electronics and- Atmospherics Research Jivision, 50 Atmosphel'ic5 Physics Secti.on, AlolOogartlo, Ne\. l;jexi.eo.

http://slidepdf.com/reader/full/contra.ct

http://slidepdf.com/reader/full/Secti.on

http://slidepdf.com/reader/full/l;jexi.eo

http://slidepdf.com/reader/full/contra.ct

http://slidepdf.com/reader/full/Secti.on

http://slidepdf.com/reader/full/l;jexi.eo



GENERAL. ELECTRIC

RESEARCH LABORATORY

Occasional Report Number 5

PROJECT CIRRUS

on

,

1. Production of Ice Crystals by the Adiabatic Expansion of Gas

n. Nucleation of Supercooled Water Clouds by Silver Iodide Smokes

m. Influence of Butyl Alcohol on Shape of Snow Crystals Formed in the-;;=:=-::::-::-7:"____Laboratory

Prepared

by

Bernard VOMegut

The researeh.re,ort.d tn this document was made posslbl. throug'

iAccesion r:or

r - - - - - - - - - - - - NTIS CRA&I oOTIC TAB

UnannouncedJllslllic:lIK"l

By - - - - - ..- . - . -_Oistribu1itv l!

A V i ! l l . ; b i l , t ~ 1 Codes---- -0--- -

'0 r ,l\o'cll and I orOlst S ~ c c : < J 1

fJ... I support extended the Research LabOlatory ofthe General Electric Com_ - -o

-.:...---'---

pany Jointly by the Signal Corps and the Navy Department (Office of NavalResearch) -under Signal Corps Contract No. W-36-039-SC-32427. Acknowledgment is also made'Of assistance from the Army Air Forces inproviding aircraft and associated personnel.

Approved by

G. E.Requisltion Schenectady, New York

EOO-2U90 15 July 1948



PRODUCTION OF ICE CRYSTALS BY THE ADIABATIC EXPANSION OF GAS

In experiments on a supercooled cloud produced

in a home freezer, V. 1. Schaefer (1) showed that

at temperatures of -38.9OC or lower water vaporspontaneously forms ice crystals in very lar(l:enumbers. By the adiabatic expansion of ai r in aWilson Cloud Chamber, B. M. Cwilong (2) found

similarly that ice crystals were produced at tem-

peratures below -350C. Simple and interesting ex -

periments can be performed by a combination ofthese two techniques.

A child's pop gun fired into a supercooled cloudin a cold chamber produces very large numbers ofice crystals. The adiabatic expansion of the ai r

as i t is released from the gun reduces it s temper-

ature to below -38.90C wlththe consequent produc-

tion of large numbers of ice cryStals. If the cork-is not put into-the gun tightly enough, the tempera-

'tures producedare above088;9OC andno ice crystals

result.

In order to rule out the _possibi11ty that the loudnoise from the pop gun might have caused nucleation,amixture of potassium chlorate and suUur was

:exploded in the supercooled cloud. Although thereport was fa r louder than the pop gun, no ice

crystals _were observed.

A bottle of carbonated beverage having sufficientpressure produces large numbers of ice crystals

when it is suddenly opened in a supercooled cloud.Bottles of carbonated drinks kept in the freezing

-compartment of a household refrigerator often be -come supercooled. Frequently, these bottles do

not freeze until the cap -is removed. If such a bottleis _watchedas the cap is removed, many ice crystals

can -be seen to form at the surface of the liquid and

spread throughout thel?ottle. A miniature snow

storm produced in the gas in the neck of the bottlestarts the crystallization of the contents. I t has

been observed that a bottle of supercooled beveragecan be caused to freeze by tapping the surface of

the conta iner. Such a tap undoubtedly causes adia-

batic compression and expansion of any minute

bubbles in the liquid which could momentarily re-

duce their temperature to below ..3S.9OC, thus

starting the formation of ice cryatals.

The vapor trails which sometimes stream offthe propeller Ups and wings of airplanes flying atlow temperatures probably are similar to the fore-

going phenomena. As the propeller or wing passer;through the air, i t causes adiabatic expansion 01

the ai r in certain regions. If the temperature of

the atmosphere is sufficiently low, this expansionwill momentarily reduce the temperature to a value

at which ice crystals form spontaneously. If theatmosphere is supersaturated with respect tOice,

these ice crystals will grow into a visible vapor

trail.

The surpl'isingly large number of ice crystals

produced by the rapid expansion of even a small

quantity of ai r can be shown by bursting a small

rubber balloon in a supercoolerl cloud. A balloonabout 1.5 mm in diameter whetl burst in a super-

cooled cloud at -200C produces at least 3by 107

snow <:rystals or about 1.6bylOl0 crystals per ccof eT.panded air.

Experiments under carp-fully controlled COIl

dit\OllS ar e being conducted in this laboratory byV. 1. Schaefer to determine quantitative relation-

ships between the number of crystals produced andthe temperature, pressure, VOlume, and humidityof the expanded air.

1. a) V. 1. Schaefer. The Production of Ice Crystals in a Cloud of Supercooled Water Droplets. SCIENCE,1 0 ~ , ~ 5 7 - ~ 5 0 , (1946).

-b) V. 1. Schaefer. The Production of Clouds Containing Supercooled Water Droplets or Ice Crystals

under Laboratory Conditions. BULLETIN OF AMERICAN METEOROLIGICAL SOCIETY, 29, No.4,

175-182, (Apri l, 1948).

2. B. Cwilong. NATURE, 155, 361-362, (1945).

2



ADJUSTABLE

COLLAR

/

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

NUCLEATION OF SUPERCOOLED WATER CLOUDS BY SILVER IODIDE SMOKES

A. INTRODUCTION along with a large volume of air into a crude wind tunnel four feet square in cross section and twenty

About a year and a hal! ago, it was found that four'feetlong. The stirring action ofthe fan andthe

silver iodide smokes had the property of causing turbulence in the tunnel mixed thp. smoke with the

snowflakes to form in a supercooled cloud.(1) It ai r to produce a more dilute smoke which was dis

believed that silver iodide particles are good nuclei charged atthe other end ofthe tunnel where samplesfor ice formation because ofthe close resemblance were taken. When a heavy white oil-smoke was In

oftheir crystal structure to that of ice. Experi troduced into the tunnel, i t appeared to be quite uni

mentation has been underway to learn more about formly mixed and diluted as it left the tunnel. The

-the production and behavior of these smokes. The rate of flow in the tunnel, as determined by measur-

work to be described in this account should be re - ing the velocity with a vane-type anemometer, was

garded as preliminary. The techniques and appa 4byl06 cm3/sec.

ratus used in the work frequently leave much to beOf the silver iodide smokes used, all except the

desired i 1 the way of precision. The results are ones having the largest particle size were com-tentative and await confirmation by better experi

pletely invisible under the conditions ofthe experiments. TI;\s work, despite the uncerhinties in it, ment. The smokes having the larges t particl es werenevertheless sheds light on the mechanism ofnu-

quite · ransparent and of a pale blue_or purple apcleation and suggests new exper iments and improved pearance in the sunlight.techniques for giving a more complete picture of thephenomena associated with nucleation by silver

-iodide. 2. Smoke Sampling and Diluting Syringe

In many oI the cases, only a fraction of a cubic-B. APPARATUS AND TECHNIQUES FOR

centimeter of the smoke from the tunnel is neededMEASUREMENTS ON SMOKES

for a .test. Precipitation on the walfs oI a container oUhis size would be very rapid . Therefore, a

-1. Wind Tunnelsyringe was constructed for taking a sample of the

-In order to determine the output of a s o u ~ c e of smoke and diluting i t quantitatively to any desired· silver iodide smoke, it is necessa·ry to secure a amount. The syringe, (Fig. 1), consis ts of a metal ~ a m p l e of smoke for testing t!lat is aknown frac- tube -three inches lit diameter with a piston and tion of the total -output. This was a c c o m p l 1 s h ~ d by leather washers which-can be- moved:back and forth -diluting the output of the smoke generator with a afixed distance. A sample of smoke was taken with-!arge known flow of aIr, and taktng a known volume of the sYringe and then diluted to the des ired concen-this dilute smoke for testing. The smoke generator tr ation by moving the piston in and out the requisite.was placed In front of a quarter horsepower electric number of times in smoke-free air, up-wind from

-Ian, three feet In diameter, which sucked the smoke the tunnel.

SMOKE SAMPLING AND DILUTING SYRINGE

FIG. 1

3



k ! m ~ ~ ~ ~ t l ~ s T A C K OF MICROSOPE

3. Cold Chamber

The early measurements on the number of nucleicontained in 3ilve r iodide smokes were made usingSchaefer's technique.(2) A measured volume of

smoke was introduced into a supercooled cloudin ahome freezer, and the nUlnber of snow crystals pro

ducedpercubic centimeter was visually estimated.

This technique has been slightly modified in themore recent work-and the apparatus used is shownin Fig. 2. The tests were carried out in a brass

cylinder 15 inches high and 12 inches in ttiameterhaving walls 1/2-inchthick to provide good thermal

conductivity •

The chamber, which was closed at the bottom, wasmaintained atalow temperaltreby placing iUn a four

cubic foot hane freezer. The freezer thermostat wasused to regulate the temperatures. A supercooledcloud was maintained in the refrigerated cylinderbyevaporating water from wet paper toweUng woundaround a 15-watt electric heater placed in the lowerpartof the cylinder. A hinged masonite lid was u s e ~ to close the top oHhe cylinder during tests. _.Thetemperature aLthe top of the cylinder was found-tobe about 20C warmer than at the bottom. The minimum temperature obtainable in the chamber was

-200c. In future experiments, it is highly desirable

thatlower temperatures be obtainable and that pro

visions be made for better temperature regulation.

Smokes were tested by introducing them from thesampling syringe into the supercooled cloud in thecylinder. A stack of cold microscope slides wasplaced in the bottom of the cylinder. The snowflakes,

produced by the action of the smoke, settled on thellottom ofthe cylinder andon the topmost micro

scope slide. At intervals, two minutes apart, thesl ide on which the snow -had fallen was removed,

thus exposing the slide beneath. The slide which wasr e m o v " ' ~ was then examined under a microscopekept 10 .. le freezer. By means of a Whipple eyepiece, the number of flakes collected per squaremillimeter in a two-minute period was counted.When the rate of snowfall had dropped to a low value,the total number of flakes which had fallen persquare millimeter was -determined by adding thenumbers which had fallen in each two-minute sam

ple. The total number of snow crystals which wouldhave been produeed by the e_Ure outputcif the gen

erator could then be calcuhited from the area of thebottom of the cylinder, t1ie -volume and dilution ofthe smoke introduced, and the volume rate of pro

duction of the smoke.

HOME FREEZER

BRASS CYLINDER

WATER VAPORIZER

THERMOMETER

- SLIDES

INSULATION BOARD

APPARATUS FOR COUNTING NUCLEI IN SMOKE

FIG. 2

4



. . -- - --_ ._ ---- - 4. Electron Microscope Examination of Smoke

Smokes being tested were examined with theelectron microscope to determine their appearanceand particle size. Samples of smoke were precipitated on a form\'ar film Eupported on a fine ";.i.rescreen by moving it for about a minute in and out ofthe smoke stream about three ftletfrom the genera

tor.The smoke stream at this point is still quitewarm. The sample screen, because i t is in thesmoke only a moment at a time, remains cool sothermal precipitation may playa part in the coHection of the smoke.

Under these conditions of precipitation, it isquUelikely that particles of a cer tain si ze may be selec

tively pl'ecipitated, so that the sample obtained is

not entirely representative of the smoke. A morereliable method would be desirable so that more

trustworthy data can be obtained.

The smoke samples were photographed using theelectron microscope. Determinationsof the particlesize and the number of particles percc of m t e r i a l were made from these photographs.

5. Smoke Generator

The smokes used in these tests were producedby a smoke generator constructed from a c o m m e r ~ cia! compressed ai r atomizing nozzle of -the so_t

used for paint spraying and humidifying (BinksNo. 174). The generator is shown in Fig. 3.pressed hydrogen gas a t 20 pounds per square inchwas applied instead of ai r to the ai r inlet ofthe Ilozzle and a solution of sUver iodide was used as ' theliquid to be sprayed. The hydrogen stream as it-lertthe ·nozzle was ignited. The heat ofthe flame vapOX:-ized the silver iodide in the spray into a gas which,

upon mixing with the atmosphere, condensed lnto :a

smoke of small silver iodide particles. In order toprevent the hydrogen flame from being blown out bythe wind, a flame holder conSisting of a piece of3/4-in. pipe two and one-half inches long was placed1/2 inch from the spray nozzle. At a pressure of20 pounds per square inch, the spray nozzle usedabout three cubic feet of hydrogen pe r minute(measured at atmospheric pressure).

6. Sllver Iodide Solutions

Although silver iodide is very insoluble in waterand organic liquids, i t is quite soluble in acetone or

w a t e solutions containing a soluble iodide such as

sodium ammonium iodide. The solutions used inthis work were made by dissolving 200 gms of AgIand 100 gms of NH41 in a mixture of 750 cc of ace

tone and 250 cc of water. A dUute solution was alsoused which was made by diluting the above solutionto ten time s it s volume with acetone. Solutions canbe diluted any desiredamount with acetone; however,

dilution with water causes precipitation of the silve r iodide. Ammonium iodide was used in the solu

tions.for these experiments because it probably iscompletely decomposed in the hydrogen flame, thusleaving a smoke of uncontaminated silver iodide.

The rate at which sUver iodide was fed into thesmoke generator was varied by controlling the rate

of flow of solution by a d j ~ s t i n g the valve on thenozzle and by using solutions of different concentrations.

C. EXPERIMENTAL RESULTS

1. Decrease in Rate of Snow Formation after

Seeding

There was found to be a large differp.nce in the

be havior ofthe supercooled cloud whenit

was seeded

HYDROGEN GAS AT 20 P.S.1.

FLAME HOLDER

~ - = ~ ~ ~ ~ ~ , = ~ ~ ~ ~ ~ ~ / ~ ~ ~ ~- .; : : ,-

- - :

SPRAY NOZZLE

SILVER IODIDE SOLUTION

SILVER IODIDE SMOKE GENERATOR

FIG. 3

5

- ' -==== ===_ :.c __ ~ .. = ==·=·= = · :... _'_o_=_o.= __



- - --- - - - - - - ----

- - ~ ~ - - - - - - - - - - - - - - - - - - - - - - - - -

with the lowtempel'ature ai r produced by:a -pop gunand when i t was seeded with sUve-r iodide smoke.From Fig. 4 it C.t.1 he seen that although-many snowcrystals were produ.cedby the low temperature fromthe pop gun, all of t1:ese crystals had precipitatedto the bottom of the container at the-end of tenminutes. When the cloud was seeded with silveriodide smoke, howevel', a measurable -number of

ice crystals were still precipitating at the end ofalmost an hour. The rate of snowfall decreases toone half each two or three minutes. T h ~ s rate ofdecrease was not found to vary signUicantly withtemperature or with the particle size 9c-the smokealthough, as it will be seen, the total number ofsnow crystals varied over several factors of ten,depending on the temperature of the supercooledcloud.

2. Precipitation of Smoke in Syringe

One possible source of error in these experiments is that which might be caused by coagulationand precipitation of the smoke inthe sa_mpling syr

inge. In order to evaluate this rate of d i s a p p e a r ~ n c e , tests were made in which small smoke samples

NEE 4z0

z" :i..J

Zi

a:UJa.

en..J

t!en

:>-

a:(, )

UJ

2

0

a:UJQJ

:E

::>z

0

2 -2(! ) 10

were withdrawn f r o ~ l lile syringe after if had:beeninthe syringe for varying periodsof time. Thenumber of effective nuclei in a sample of smokeowaEfound to decrease by one half every twenty m i n u t e ~The time required to take, dllute, and discharge asample into the cold chamber was never more thana minute or two so that the changes in the smokeoccurring d u r i n ~ this time were not large.

3. Number of Ice Crystals perGramof Silver Iodide

The results of experiments carried out with dif

ferent settings of the smoke generator and in supercooled clouds at different temperatures is shown inFig. 5. In this graph, the results are also given forthe electron microscope examination. The number·of particles of silver iodide per gram of silveriodide was computed by measuring the approximate mass of silver iodIde on a given area of thesample and then counting the number of particles inthe same area. The computation was based -on theassumption that the density of the particles was-thesame as that of silver iodide crystals. Figure 6

shows electron photomicrographs of a typicalsmoke.

SUPERCOOLED CLOUD AT -17·C

SEEDED WITH POP GUN

SUPER C_OOLED CLOUD AT -17°C SEEDED

SILVER IODIDE SMOKE

Cl

0

-I

I' 40 50 60

0..J

FIG. 4

6

="



ERRATA

The calculations of the number of particles per gram of slIver iodide

from -the electron photomicrographs were based on determinations from the

photographs of the averags dl3 .meter with respect to volume. I t has been dis-

-. covered that this qnantity, computed as

i . 'nd4

£nd3

,where·n is the number of particles and d is their diam.eter, is not properly ap-

.plicable to these calculations. In view of this mistake, the data from the photo

' g r a p h ~ L h a v e been re-calculated properly. A hundred or more particles on

" representative are.as of each of the photographs were measured and counted.

·From this data, the number of particles per gram of silver iouide was computed

'by dividing the number of particles in a given area by the total mass of the parti-

" ·cles in that area computed on the assumption that they were spheres and that

they had the density of silver iodide crystals.

The smallest particle which can be seen with the electron microscope

o -' ,

ip of the order of 3.0 A diameter. In al l of the photographs, a large percentage

'of the pa'rticles were of this size so that it is probable that many particles in

the smoke are too small to be seen. The calculated numbers of particles per

'gram is based only on those particles which were visible, so that they can be" <

..

'. takep ·as minimum values.

The new values for the number of particles per gram are in al l cases

<

. 'higherthan .the number of snowflakes produced per gram which is what would

' be expected as an of the particles probably do not serve as nuclei. The corrected

'curve i s h o w n 'in· revised Fig. 5.

re'



-- ' - . - ~ ~ - 5 3 0 A ~ ' i ~ ~ & ~ O F = - ; : A P A R ~ ~ ~ ~ O & 3 0 _ 1 1 0 0 A 19 . .. I .. . ... .... .... .. . . .. ....... . .. I I' : " ~ ~ - - - - ..... ' ' .. 1 8 ~ - - - - : - ____ ,"

' / " .., ,~ - - - - - -r 7 ~ 'NUMBER OF ,SMOKE pARtiCLES P ~ ; ~ A ~ + - - - ~ - - ~ ~ - - - o - - . O E T E R M I N ~ b ,FROM" EL.'ECTRON PHOTOMIC 'ROGRAPHS . --- __ '

1 6 ' , ", -0' . " , ---

-20 0 e15

14

~ 1 3 ° C 13

'

1 2 ~ - .;.IO°C - - - --

II"

10

9

8

7

6 NUMBER OF ICE CRYSTALS 'FORMED PER GM. OF AO I RELATED TO TEM.PERATURE ,OF SUPER COOLED

5 LOUD AND RATE OF AgI CONSUMPTION.

4

-3

('REVISEO) FIG.52

.I .2 .3 ' !I- .5 .7 I 2 3'4 5 7 10 20 30 40 50

MG/SEC OFAgI CONSUMED BY SMOKE GENERATOR



- - - - - ------

SIZE RANGE OF SMOKE -PARTICLES

130-1400 A18

17 ____ - o L ~ ~ B E R OF SMOKE PARTICLES DETERMINED FROM PHOTOMICROGRAPHS...GO

16 _( - ~ - - - - ---0-._---"----

15 - v - - .. . _ -----0--

14C)

- ~ - - - - - - - - - - - - - - - - - ~ i: 13

Ie12

en

en "5

9

80

a: 7UJCD

2 6 , -NUMBER OF ICE CRYSTALS cFORMED PER GM. OF AgI-RELATED TO TEMPERATURE OF SUPER COOLEDC ~ O U D AND RATE OF AgI, 'CONSUMPTIONi

-j '

§5;-

-MG/SEC OF AgI CONSUME-C! - -BY SMOKE GENERATOR

FIG. 5

In one:region the data indicates that the number trates a s1gi1l!icantd1fferencebetweenthc behavioroUce crYStals which form is greater than the num c l o u d seeded with sUver :lodide andone seededbe r of sUyer-iodide particles iiltrOduced. Probably , - P 9 P g u n . Seedingwltha popgUn or dry ice pro-

this is f h e result of experimen1!t1 inaccuracies. c ~ s very large numbers oLsmall ice cryStals byThese errors could have their source in deductions cOQI!nga small region ofthe :cloudto a temperature

made -from the electron microsc.ope data or in the at -whJch Rpontaneous nucleation takes place. Thesesampling 'and ice crystal counting -technique. It ice:crystalsthen mix with _he cloud and grow at theseems doubtful under the conditions of the experi expense of the supercooled-wafer drops. From Fig.

ments tliat one particle could cause more than one 4, i_ can be seen that although-a'large number of icesnow"aJCe-,to form. - crystals is produced by gun, al l of these

crystals have prec ipita ted out at "the end of ten min

D. DISCUSSION OF RESULTS utes; Now if al l of the sUveriodlde particles putThe-workthusfar is not o! s u f f i ~ i ~ n ~ scope to glve t n t ~ j h e cloud formed ice crystals at the time they

strafghtfC?rward answers to many_questi ons which we-e.:introduced, they, too, should have preCipitated

arise- concerning tbe mechanism- of sUver iodide ouh.t the endoften minutes. However, at theenrl of-n u c l e . a t 1 Q . ~ . It is the purpose i l . ~ ~ ~ s c u s s i o n to almo_t an:hour , ice crystals are stlll.precipitatlng.veniureosome possible explanations.for the experi Thiscleads one to the conclusion-that al l of the sU

mental q ~ s e r v a t i o n s ver :iodJ.de parUclesdo-notfornl'ice crystals imme

diate-y and that even at the end of hali an hour or

1. Formation of Ice Crystals more, appreciable n u m b e r s s l l v e r iodide parti

F i g u r ~ _4,-whlch shows the n t i m . t > ~ r of snow crys cle-shave notyetformed snow_cryStals and are still

tals p r e c ~ p 1 t a t i n g per minute a f t ~ r seeding, Ulus- present in the cloud.

7



. There are various possible explanations for thetime required for silver iodide particlesto initiatethe formation of ice crystals in a supercooled cloud.One possibility is thatin orderfor-an ice crystal toform ona crystal of silver iodide, a cer tain crit icalnumber of water molecules must by ~ h a n c e arrangethemselves on it s surface in the structure of ice.Accordingtothis explanation, a silver iodide parti

cle would not act as a nucleus until this event tookplace. The presence of a silver iodide surface mightbe regardedas merely greatly increasing the probability of the formation of ice.

If we take a simplified view of the theory of nucleationas advanced by Glbbs, in supercooiedwater or

In a region supersaturated with respect to ice, thewater molecules by chance,from time to time, ar-range themselves in the lattice of crystalline ice.I f these minute aggregations are small er than a cer-

tain Size, their vapor p ressure is greater than thatof the supersaturated region and they are unstableand break up. If , on the other hand, they are larger

than a certain size , their vapor pressure is such that

they are-stable and continue to grqw. Thp.lower thetemperature, the smaller will be -the crItical sizenecessary for stability. Schaefer (3) has found,using clean ai r free of dust, thantie rate at whichnuclei occur is -very small at t e ~ p e r a t u r e s above

...

. '..' -'

, •• ! .& •

. ,.... .to • -. '_e .....

.. ..

-3S.DOC. However, in the presence of a si lver iodidesurface, i t is possible, because of the close similarity between ice and silver iodide, that the probabillty ofthe chance formation of a nucleus is greatlyincreased and is large even at temperatures ashighas -IOoe.

Another way of looking at the phenomenon is toconsider the growth of an ice crystal as the formaaUon of a crystal of ice on ice. We know that thistakes place with the greatest ease. The formationof ice on a large surface of ice requires the formation of very little new ice surface and hence thereis little change in surface energy. For every newice surface formed, an almost equal ice surface Iscovered up. However, when a new ic e surface is

formed in the absence of any other surface, largeamounts of surface energy are required reiative tothe free energy decrease in the formationofthe interior of the new phase .Because of the close simi-

larity-of ice and silver iodide, the formation of iceon a silver iodide surface:probably involves only -asmall amount of surface energy, and therefore, the

the-chances of thist a k i n g p ! ~ c e

are good.The lower the temperature, the smaller will be

the:critical size of a stable nucleus. Because t h ~ cr!tical size is small atlow temperatures, thechances of a nucleuS forming are, in genf;!ral, better

FIG. 6 ELECTRON MICROGRA.PH OF SMOKE PRODUCED AT 2.3 MG/SEC AGI CONSUMPTION

8

http://slidepdf.com/reader/full/MICROGRA.PH

http://slidepdf.com/reader/full/MICROGRA.PH



Hence, the rate of-nucleus formation is generally probably not the most important rate-governinggreater at low t e ~ t l p e r a t u r e s . factor.

Another possible explanation f o ~ the time re

quired for silver iodide particles toform ice crys

tals can be based on the assumption tl13.t the silveriodide particles produced by the generator are notall in the hexagonal form which is similar to ice.

Possibly they might exist in a metastable conditionas a supercooled liquid or as some other modification. In this case, thp. r a t ~ at which they reactto form ice crystals ~ I o u l d be determined by therate at which they transform into the stable h ~ x a -gonal form. However, it appears for several reasonsthat this factor is probably not large in these exp e r i m ~ n t s . It has been found that the silver Iodidesmokes used in these tests will nucleate silveriodide solutions supersaturated with -respectto t.hl!

hexagonal form and that the crystals which resultare of the hexagonal variety. 'fhe number of suchnuclei in a given volume of smoke is of the sameorder of magnitude:asthe number oLparticles determined by the electron microscope. It, therefore,

seems probable .that at room temperature most ofthe smoke par ticles are oUhe hexagonal structure .In addition, one would expect, i f the rate-dete rminlngfactor involved the transformation of the sUveriodide par tic les Uiemselves, that smokes aged fora-period of time at -low temperature would show adifferent rate of falling off of ice crystalformaUon·when they are puUn a supercooled cloud. Thisbasn.ot-been observed:

Another possibility is that the formation of anice .crystal under the influence of !l silver iodideparticle can take place only in the liquid phase. In

Hiis-case, the rate of ice r:rystal formation mightbe limited by the rate at which silver iodide par

ticles entered into supercooled water drops. Thiscould occur either by diffusion cifthe:particle to thedrop or by condensation of a drop on 'a particle. It

has been found that when a suspension of silveriodide particles In water is sprayed: into a supercooled cloud, ice crystals are formed. I t is, therefore, probably that -silver iodide partic les can actas nucleating agents in a liquid water drop. However,it _has not been definitely established-that this is anecessary condition for silver iodide· o act. If weassume this to be the rate governing factor in thesee_xperl.r.lents, we are faced with the :problem of explaining the large:effect of temperature -on the total.l)umber of ice crysta ls produced. As will be dis

cussed later, the v a r ~ a t i o n of the number of ice

crysta ls produced with temperature can be explainedon the bas is that the of nucleation increasesgreatly with decreaSing temperatures. Th(re is noreason to expect that t h ~ rate at which particlesenter water drops is particularly' temperaturesensitive. It is quite possible that the time requiredfor a particle to enter a water drop is of considerable importance in these experiments but it is

The interpretation of results is further complicated by the possibility that the small particles ofsilver iodide smoke may dissolve in water dropsbefore they have a chance to start the formation ofan ice crystal. At 250C the solubility of silver

iodide In water is about IbylO-8 mols/uter'(4) Awater drop 10 microns in diameter on this basisshould be capable of dh,solving a silver iodide par

ticle about 100 .R in l ~ l a m e t e r . The solubility shouldbe somewhatgreater for particles of this small size.However, the effect of particle size should becounteracted to some extent by a decreased solubility at the lower temperatures. If these factorsare of Importance, the kinetics of the nucleationmay be dependent to a large extent on the rate of

solution of the silver iodide particles in the waterdrops.

The author is inclined to favor the explanationthat the rate of ice crystal formation is governed

primarily by the rate of spontaneous ice nucleusformation on the;silver iodide particles. In someas yet u n r e p o r t e ~ eX!;leriments on _he nw::leation ofsupercooled water and supercooled . in, the authormeasured at con.stant t e m p e r a t u r ~ the rate of solidification of systems c o m p o ~ d of a number ofindependent s u p ~ r c o o l e d drops. I t was observedthat the drops did not all freeze once, but frozeat a rate which ~ e a d i l y decreased-with time. Thisrate of freezing-greatly increased as the temperature was lower_d. In these experiments, foreignparticles and surfaces were pres_nt which undoubtedly served as foreign nuclei. The time requiredfor these drops' to freeze could be best explained:on the basis of the chance formation of stable nuclei.

on the foreigl .urfaces. It seems reasonable tobelieve that in the case of nucleation by silver iodidethat similar phenomena playa dominant role.

If we proceed'on the assumpUonthat si lver iodidemerely increases the rate of the chance formationof spontaneous-ice crystal s, it is ,possible to cometo some conclusions as to the magnitude of the rate'and it s dependence on temperature. The particlesof silver iodide' introduced into the cold chamberundoubtedly become lost as potential nuclei by pre

cipitation on the walls of the chamber. The rateofprecipitation on the walls is probably large becauseof thermal diffusion resulting from the 1 5 - w a t ~ heaterin the vaporizer . One would expectthe rate at which

particles disappear by precipitation onthe walls tobe proportional tothe concentration of the particles.On the basis of this assumption, the concentrationof particles would decrease exponentially with time.If the rate of sno.w formation is proportional to theconcentration, this, too, should ·decrease expo..,nentially with time. This is experimentally observedto be the case.

9



On the basis of the foregoing assumptions, onecan analyze the situation mathematically.

I£ c is the number of silver iodide particles per

-unit volume at a time t then:

(1)

where K1 is the rate at which the silver iodide par-

ticles form ice crystals and K2 isthe-rate at which-they ar e being removed from the cloud by precipitation or other causes. By- integrating Eq. 1, weobtain an expression for the concentration c at anytime:

(2)

where Co Is the concentratirii at time t =0 . I£ N is

the number oLice crystal:, formed per unit volumethen,

dN _ 'K - K :e·(K1 + K2) t (3)dt - - 1 • c - l ~ O

:which when integrated gives for the number of'crystals fornied at time t:

N 1 - e-(K1+ K2)t1c6 (4)Ki_+ K2

-Or when precipitat ion is complete

K1 ••. '(5)

N = Kl + • Co

or

(6)

The fact that the rate at 7o1hich the rate s n o w f a l l -decreases with time is not -greatly Influenced bytemperature, -while the total ' number of- crystals

_produced increases by a large factor with a small:temperature ~ e c r e a s e , indic:ites that at h 1 g h ~ r tem·

peratures K2 i s fa r greater than K1. This!s anotherway of saying t h a ~ at higher temperatures-the rate

at whichthe silver iodide particles form snowflakes.is so small that a large majority of them precipi tateon the container walls before they have a ~ h a n c e toform snow crystals. At -200C the number-of snow-crystals o b t a ~ e d is about Uie -same as thenumber

of particles -determined f r o ~ the c l c c t r o n m i c r o ~ scope so tha t:it is reasonable to assume that atthis

-temperature_practically all oUhe smoke particles

form s n o w f l ~ e s . If we assume that the ra t_ at which precipitation

decreases wit!l time at the 'higher temperature is

determined solely by K2' then K2 computed from-the curve in Fig. 4 Is approximately 6bylO-3 pe r

sec. From the data in Fig. 5, N1 and Co can bee ~ i m a t e d and from them and the above value for

~ 2 ' values for K1 for dUferent smokes atdUferent-temperatures can be computed.

These values expressed a shaU life in hours ar e

as follows:

Mg Agi/Sec

1.5

Range ofParticle Diameter

30 - 700R

Half Lifein Hours

.130C ,.IOce5.5 17S

7.5 50 - 1070K 1.9 7537.5 130· 1400K 1.4 67

If this data is extrapolated on the assumptionthat the log of the nucleation changes linea rly withthe reciprocal of the absolute temperature, we findthat at ·5 0C the haU life is several years while at

~ 2 0 0 C i t is a few seconds and at ·2SOC a few milli-second:;. '

This large effect oftemperature on the nucleat ionrate is of the same order of magnitude as that ob~ r v e d in the_measurements on -supercooled water

drops and on superc(:olecl tin drops.

The curves in Fig. 5 show a maximum for thenumber of nuclei produced per_gram of silver iodideaJ a certain rate -of silver iodide consumption. This

!,s probably because at a low rate of cons!JDlption,the particles produced are s m a l l that their re -

a,ction rate is small and many 'of them precipitateout before they--have a chaJice -- to serve as nuclei.-At high rates of-consumption, ;the particles formedare larger' and react more rapidly,but b e c a u ~ theyare larger, fewer ar e formed per gram of -silver

~ q d i d e .

2. Diffusion C<:>efficient of Smoke

An average -' diffusion coefficient for the silver

iodide smoke can-be c a l c u I a t e d ~ r o m measurements-made on the rate at which the smoke p r e c ~ p i t a t e son the walls of a container. In the experi ments which'l!ere made to find the rate at which file numberofnuclei falls off :as a fWlction of-the tiOie the -smokeis held in the sampling syrirlge, it was fOWlt:l thatthe hail life ofUle smoke ~ o u t twenty minutes.This was for a ' smoke produced:when the g e ~ ~ r a t o r was consuming'20 mg/sec of1.gI.

Langmuir bas derivE.ll the :following e ~ : p r e s s i o n !cwthe rate of:precipttatlon-Qf a smoke of particlediameter A on -tIle walls of a- container having avolume V :tnd!!n intern2! 2rea A with Blight-convection caused by a siightly h!gher temperature at~ h e bottom than-the top.

d 1n c 0.64 A D2/3

(7)( i t = V

10

http://slidepdf.com/reader/full/derivE.ll

http://slidepdf.com/reader/full/derivE.ll



I

r

where· c is the concentration of the smoke at any

time t.

Using this equation the diffusion coefficient is

calculated to be approximately 6bylO-5.

Langmuir has also given the following expressionfortl:!edUfuBion coefficient as afunction of particle

radius:

D =-16

2.04X210 +

1.18 x 10-11a (8)

a

where a is the particle radius.

According to this expression, the diffuslon·co

efficient 6by10-5 corresponds to a particle diameterof 400R. This is a reasonable agreement with themeasuremen ts from the elec tron microscope whichshow this smoke to have ·a particle diameter rangingfrom 100 to 1400 with a median diameter of

about 300

As ·has been shown K2' _or the rate of disappearance in the cold chamber -of Silver iodide particles

from:causes other thal! n u c l ~ a t i o n , is about 5;7by10-3:Per osecond.

Using the diffusion c ~ ! f i c i e n t .of 6by10-5 -andEq. (7), -it can be calculated that · the smoke concentratio-n should fall at a rate of 1.3by10-4-,per

second. This is far less -·Ulan the obserVed rateo!

d e c r e ~ s e , K2, which via_ -found to be 5.7bylO'::3.Thls 'calculation, of course, does not take-into,ac

c o u n t t h e thermal d1fiusion:caused by the vaporiZerheater In the coldchamber'whlchwould be e x p e ~ ~ e d to greatly increase the r ~ t e of precipitation' on thewalls.

Using.the ciUfusion c o e ~ £ i c i e n t i t is poSSible -to

estimate the rate at whic·h-the concentration of, thesmoke ·decreases because -of precipitation ol! , the

water -drops of t r . ~ :.apercooled cloud. The rat.e.of c h a n g ~ of concr:ntration ·d!1e to this cause is givenby the:relation

d 1n c"""""iif"""" = -4fT D fO Q (9)

where rO is the .radius of the water drops and Q is

the number of them per cubic centimet er.

I f we assume the liquid water ·content to be 1 gm

percubic me.ter and the drot> radius tobe 5m1crons,the rate of change of ~ o n c e n t r a t i o n from this causeis 3.6byl0-4• .

E. SUMMARY

Measurements have been carried out on the nucleation of supercooled water clouds by silver iodidesmokes. The·smokes were produced by spraying asolution of silver iodide into ahydrogenflame. Theparticles of the smokes were found by electronmicroscope examination to range 'in diameter from

30R to 1400 R. The number of ice crystals pro

duced per gram of silver iodide -was determined as

a function of the temperab:re of the supercooledcloud and the·rate of Introduction of silver iodideinto the flame. Yields of ice nuclei of as high as

10 16 per_gram of silver iodide were obtained whenthe s u p e r c o ~ l e d cloud was at - 2 0 ~ C . At -10oe, thesame smoke produced only 1012 :7 ice crystals per

gram.

It has been found that silver - iodide particles do

not react ~ m e d i a t e l y to form ~ c e crystals "Nhen

they arepu_into a supercooled cloud. Ice crystals

were found i st1ll tobe forming at;i1;measurable rate

fifty minutes after a silver iodide smoke was in

troduced-into a supercooled cloua. It is believedthatfewerice-crystals are produced at highertem

peratures than.at lower temperatures because thesilver iodide_particles react more slowly to formice crystals-and most of them· precipitate on the

walls of the.c9ld chlUIlber before:they have time toreact.

According' to this interpretati0 ll of the results,

the rate ofre.action at-130C istli(rty. or forty times

that at -10°C.

11

http://slidepdf.com/reader/full/rat.e.of






INFLUENCE OF BUTYL ALCOHOL ON SHAPE OF SNOW CRYSTALS FORMED IN THE LABORATORY

In the course of laboratory measurements on thenumber of ice-forming nuclei containert in varioussmokes, a microscope was set up In a refrigeratadbox for the purpose of counting snowflakes. A super

cooled cloud was formed in the refrigerated boxat -20OC by the Schaefer technique (1). Smoke containing silver iodide nuclei was introduced into thecloud and the resulting snow crystals which formedwere allowed to fall on a slide where they wereexamined under a microscope. The crystals thusproduced were predominantly in the fOl'm of flathexagonal plates.

Without any intentional change in the experimental set up, it was noticed that the type of snowflakes produced had changed from the hexagonalplates to hexagonal prisn1s having a length of theorder of five times their diameter. I t was foundthat hexagonal prisms were produced until the ai r

in the box had been cleaned out by displacing it withai r from the compressed ai r line. When this wasdone, the flakes formed were once more hexagonalplates. The cause for this change in the shape ofthe crystals was finally traced to the presence of asmall amount of normal butyl alcohol vapor in thelaboratory atmosphere which had resulted from ac-

Cidentally spilling some Qf this liquid.

The modification of crystal shape caused bytraces of butyl alcohol vapor was found to varyconsiderably with its concentration in the air in thecold-chamber. When the partial pressure of the-butyl alcohol was of the order of 10-6 atmosphereor less, no effect was noticeable. At a partial. pres-

sure of the order of 10-

5

atmosphere, the longprisms were formed. At still higher partial pres--sures, the effect diminished and hexagonal platesformed once more. The effect cf the butyl alcoholvapor on the crystals was found to be similarwhether the cloud was seeded by silver iodide smokeorby'passing apiece of solid carbon dioxide throughit.

The modification of habit produced in the pres-

enceofbutylalcohol1s similar to the changes which,

have been r ~ p o r t e d in the habit of crystals grownfrom ~ l u t i o n s to which various substances havebeen aClded, For example, scdium chloride, whichusually crys talli zes as cubes from aqueous solution,

forms octahedra if urea is added to the solution.The suggested explanation for this change of c r y s t ~ l habit is thatabsorption on the crystaUaces changestheir relative rates of growth thus modifying theirshape.

It seems probable that butyl alcohol alters theshane of the snow crystals in a similar manner.Apparently atvery low concentrations olthe vapor,the effect on the rate of growth is small on all faces.However, as the vapor concentration is increased,a pointis reached at which the rate of growth of thesides of the prisms Is greatly reduced relative tothe rate of growth orthe prisms' ends. At highervapor concentrations, apparently the absorption on

the v a r i o u ~ faces becomes more equal- so · that"theeffect on the rate of growth of the faces -becomesmore nearly equal, and the effect on the shape ofthe crystal is not very large.

Isobutyl alcohol and allyl alcohol have beenfouridto have an effect similar to butyJ alcohol, andJUspresumed other higher alcohols would -behave-in asimilar fashion. Ethyl alcohol did not show -the effect and, i f anything, seemed to favor formatIon ofhexagonal plates,

V. 1. Schaefer of this laboratory has -extendedthe scope of these experiments and this work:wtllsoon appear in publication.

ACKNOWLEDGEMENTS

The author wishes to thank Dr. Irving Langmuirand Dr. Vincent J. Schaefer for their very helpfulcounsel and suggestions in this work. He g r a t ~ -ful to Mr. E. Fullam for making the electrqnmicroscope examinations and photographs .Mr. Kiah Maynard and Mr. Duncan Blanchardcforinvaluable help in testing the smokes.

1. '1'he Production of Ice Crysta ls in a Cloud of S u p ~ r c o o l e d Wat.er Droplets. V.I. Schaefer. SCIENCE, 104,pp. 457-459, (1946).

REFERENCES

1. V o n n e ~ u t , B. Journal of Applied Physics, 18, No.7, pp. 593-595... July, 1947.

-2. Schaefer, V. J. Science, 104, pp. 457-459, November 15, 1946.

3. Schaefer, V. J. Bulletin of American Meteorological Society, 29, No.4, pp. 175-182, (194S).

4. Hill, A. E. lACS, 30, p. 68, (190S).

12



. ' - - - . . - - .. _-- _ ._ ._- --------_._-- ---- - - --- -

DIS'rRIBUTION LIST FOR REPORTS ON PROJECT CIRRUS

No. of . Copies

50 Transportation Officer,· S C E L ~ Evans Signal Laboratory, Bldg. 42,Belmar, New Jerself' Marked: "For Signal Property Officer." .

50 Chief of Naval Research, Navy Department, Washington 25, D. C.Attn: Code N428

50 Chief of Staff, U. S. Air Force, Washington 25, D. C. Attn: Dir. of R D, DCS/M

1 Director, Mt . Washington Observatory

2 Divinity Avenue, Cambridge 38,.Massachusetts

1 Committee on : G ~ o p h y s i c a l SCienf'es, Research and"DevelopmentBoard, Washington 25, D. C.

1 Panel on Meteorology, Research and Development Board, Washington25, D. C.

1 U. S. Weather Bureau, 24th and:·M Streets, N.W., 'Washington 25, C.A ttn: Dr. H. WeXler

1 U. S. Weather Bureau, 24th and ':rv1 Streets, N . W " , . · W ~ s h i n g t o n 25, C.

Attn: Dr. Ross-Gurm

1 -Director, National Bureau of S t ~ d a r d s , Washington 25, D. C.Attn: Mr. Hugh"Odishaw

1" -Department of Agriculture, Washington, D. C.

2 'Commissioner, ·Bureau of Reclamation, Washin€ton 25, D. C.Attn:: Section 724

1 Chief Hydraulic ·Engineer, U. S. :Geological Survey, Washington' 25, D. C.

Attn: Div . ofWater Utilization .

3 Director of AerQnautical Reseal'ch, N.A.C.A., 1 F Street, N.W·., .Washington, D. q.

r N.A.C.A. Laboratories, Cleveland Airport , Cleveland, Ohio·Attn: Mr. L. A. :Rodert

1 'Massachusetts Ipstitute of Techpology, Department of Meteorolog·y,Qa.mbridge 39, :MassachusettsAttn: Dr. H. G·. Houghton



- - - - - - - - - - - - - - - ~ - - - ... '- .. ~ . - - - - - - - - ~ ~ .'No... Copies - - - - - - - - - - - - - - - - - - - - - - - - ~ ~ - - - - - - - -

1 New York University, Department of M(Aeorclogy, New York, New YorkAttn: Dr. B. Haurwitz

1 New York University, College of Engineering, New York, New YorkAttn:. Dr. A. F. Spilhaus

1 University of Chicago, Department of Meteorology, Chicago, IllinoisAttn: Dr. Horace Byers

1 University of California at Los Angeles, Department of Meteorology,Los Angeles, California, Attn: Dr. M. Neiburge"C'

1 Pennsylvania State College, Division of Meteor·)logy, State College,Pennsylvania, Attn: Dr. H. Neiburger

1 Director, Blue Hill Observatory, Milton 86, Ma5sachusetts

1 Institute for Advanced Study, Princeton, New J'ersey, Attn: Dr. J. VonAttn: Dr. J. Von Neumann

. 1 Woods Hole Oceanographic Institt),tion, Woods Hole, MassachusettsAttn: Dr. C. O'D. Iselin

1 St. Louis University, 3621 Olive street, St. Louis 8, MissouriAttn: Dr. J. B. Macelwane, S. J.

1 University of Texas, Austin, Texa.s, Attn: W. E. Gordon

1 New ,Mexico School of Mines, Box 6000, Station A, Alhuqueque, NewMexico, Attn: Dr. E. J. 'Workman

1 University of New Mexico, Albuquerque, New Mexico, Attn: Dr. V. H. Regene

1 Scripps Institution of Oceanography, La Jolla, California, Attn: D. Leipper

1 Stanford University, pato Alto, California, Attn: Dr. S. Chapman

1 University of Alaska, College, Alaska, Attn: Dr. K F. George

1 Chief.,. state '}later Survey Dtvision, Urbana) Illinois

1 Landipg Aids Experiment station, Transocean Air Lines, Inc., Arcata,Califorp.ia, Attn: R. Champiol1

3 Project Cirrus, General Electric Flight Test Hangar, Schenectady County Airport, Schenectady, New York, Attn: Comdr. E. B. Faust

10 Commercial Service Section, Syracuse, New York, Attn: C. 'Po R e y n o ~ d s

65 Reser.ve C ~ ( ) r r ( \ r r A

http://slidepdf.com/reader/full/Califorp.ia

http://slidepdf.com/reader/full/Reser.ve

http://slidepdf.com/reader/full/Califorp.ia

http://slidepdf.com/reader/full/Reser.ve




[ This page is intentionally left blank. ]




Distributed By D TieInformation For The Defense Community

Documents

Project Cirrus Progress Report No. 5