£«
ih.
■h^;
THE VEIOCTTY \'''V '^'VTIO _e_ FOR TtlE
m
PRI"ARY MW ?ECONr/VRY 3 RAYS
OF •R'.rTil''.
BY
S.J. ALLEN.
DISSERTATION
SUBVITTED TO THE BOA.RD OF UNIVERSITY STUTIES
OF THE JOHNS HOPKINS UNIVERSITY
IN CO'TOR"ITY WITH THE REOUIRE"ENTS FOR THE
TEGREE OF DOCTOR OF PHILOSOPHY.
BALTIMORE
1906.
r^iroL
THE YEIOCITY \'ir RATIO _^ FOR THE'
PRI^'ARY AND SECONDARY P RAYS
OF R'\ri'.r".
3Y
S.J. ALl.KN.
BIOGRAPHY.
Sa"i.uel James "aclr.toyh "lien was born at "aitland,
"Jova rcotia on October 5th, 1877. Hia earlier educatjon
was r.ainod in the public and private schools, and in the
Halifax High rchool from whicn ha rraduated in 189 5. In
IB^G he entered the depirtnent of Electrical Engineering
at "c^ill 'Jni versi tjr , "ortreal, Canada, and graduated
witi tne de-^ree of B . r- c . in 1.^00. On rTaduation ho .vas
appointed demonstrator in the Engine eri nrr laboratories,
where he ras also enraged in research vTork on "The Plow of
"Vater throurh Pipes 'ind Hends", for 'Yni ch in ijOl he vas
awarded the derroe of ''.."vc. He was then appointed demon-
strator in the '^hy;-ical laboratory at the same university,
w:iere he re"iained for the next two years engap^ed durinr
a great part of the time in researches on "The Radioactivity
of the'i.t mo sphere", both in conns ct ion with Professor
Rutherford and alone. The result of these Inve sti -ation s
has been published in various articles in the Phil, "a?.,
Phys. Review, Phys. Zeit., and other journals.
In 1903 he entered the graduate department of Physics
in the Johns Hopkins University , obtaining the Fellowship
in ^hysics for the rear 1903-01:. Turing his course at
this Hniversit" n p has st'aried under Professor Ames, Pro-
fessor IVood, Professor Jones and Professor "orley- his
principal subject being Physicn, first subordinate "Physi-
cal Che'^.i^tr^' and srecond subordinate ''at henati cs .
''r. Mien is a "ie"iber of the Canadian ^ocietv of
Civil Engineers and the ^r.srican Phvr;.ical "ociety.
INTRODUCTION.
The reasons which led the author to : n v e s t i "■ a t e this
subject ought first to bo briefly stated. V/n.ile trying
to denonstrate experinmntal ly the iiagnetic and electro-
static devia+jon of the 3 rays from radiun, by means of
the electrical '-.etnod, I was surprised to find that no
appreciable electrostatic deflection could bo obtained
by tne methods ax first used, and that moreover tne mag-
netic deflection observed v/ a s rn u c n less than '^ e i.v o u 1 d
be led to pxpect frora the results of ether e xper iraenter s •
Considerinp: the great importance of this subject, affording
as it does the only metnod ffn nave for testing the be-
haviour of the electron at speeds approacning that of
light, it was deeded advisable to nuke a thorough inves-
tir,at:'on, and to find out, if possible, \\rherein lay
the difficulties, and the non-success of zae early
experiments.
Before describing ny own experiments it will be
of interest to y'ive a brief resume of the \^/ork done by
previous investigators on this subject, and the conclu-
1
sions reached by thera. Becquerel v/as the first to state
1 . C". V. " 13 O", ' 1 9 f'.".
that the P rays fron rad'^un vrere deflected in an elec-
trojjtatic field; wad by "usas <r i nf; this deflection,
and knowing frcn earlier experiinents the deflection of
the sane rays in a magnetic field, he was able to cal-
culate their velocity, and the ratio ^. e being the
charge in ele ctro otati c C.G.S. units, und 21 tae mass
expressed in grammes.
His experi.Tiental method aight be briefly described
as follows. A. pencil of 3 rays v/as allo^sd to pass
upward bet\-reon two parallel brass plates, and to fall
upon a photogra[>hic film, placed horizontally above
t>.em. The plates were insulated from one another, and
maintained at a high difference of potential by means of
an electrostatic machine. A tnin slieet of mica v/as
placed vertically in tie path of the rays, so as to
cut it into two nalves. ihe image on the photographic
film would thus l-e divided into two parts by a fins
line in the ce-'tre. If now on ttie application of an
electrostatic field the rays are bent, then there
ought to appear on the image a shadow, the widtli of
v/i-:ich ./ould co-^respr,nd to the least deviable ray.
Becquerel observed such a shadow, and measuring
the wid+ (f it, ■ind knowing from other e x;^er iment s
the value of the t". a r r. e t i c deflection, hu calculated
10
tne velccitv of the P particle to be about 1.6 x 30
7
ens. per sec., and the ratio je_ :i s 1 x 1 . T li i s
fn
experi.r. ent of Becquerel was "■iude in air at atnospheric
pressure, and for this reaeon no very accurate esti-^ate
of tiie true sl:ctric field betv/een the plates could
be -".ads, sinue t::e ionization cauted by the B rays v/ould
to a great extent disturb the potential gradient.
The results of iiecquerel can therefore be only con-
sidered as ap'iro xi nat e .
1
The "lethod used by Kaufmann to obtain the velocities
of the ,3 rays was entire I37 different and .vas based on
the principle of crossed spectra. The hetero2;eneous
pencil of rays from a s'lall speck of radium passed
upv/ard ijetv^een two brass plates, through a small hole
in a netal diaphragm, placed !icr i zontal] j' above the
plates, and then fell normally upon a photographic
plate, wrapped in a thin envelope of aluminium. The
brassplates were insulated froxn one another and could
be kept charged to a high difference of potential by
.Tieans of a battery of small lead ac curriul^.tor s . The
whole apparatus v/as enclosed inside a glass vessel from
which the air could be exhausted. A -na:^netic fisld was
1. Nachr. d. (".esT d . V/'ei's'ii" m" G t't'. ", ~ 1 9 iT [
applied parallel tc the electrostatic and an exposure
given for a certain time; tile c;irecti.on of the electrc
static field was then reversed, -intl i further expos-
ure of the fja-ne time as before piiven, the direct:ion
of the magnetic field beinr tho same in both cases.
'iVhen trie pls-te •vas developed there a-^'Oeared on it
t'vo curveri lines, snovving that the (3 particles had
been deflected by both the m.a,7netic and electrostatic
fields. By measuring the deflections obs8""ed, the
value of the velocity, and ^ for each rav could be
calculated. lie found that the velocities varied from
10 lO
2.36 xlO to 2.85 xlO , and the value of €^ from
7 7 )n
1.31 X IC to .63 X 10 . Assuming; that the charfre is
constant, these results showed that the mass of the
electron apparently increased as the velocity of
light was approached.
A theory nas been developed by Thomsonjand elab-
orated b^r Heaviside, Abraham, and o t her s - '.'/hereby the
mass of the elsctrojj is considered as entirely elec-
trical in its nature, and increases with the speed
of the electron, reachin,?^ at the velocity of lifht
an infinite value. The late values of ^ obtained
m
by Kaufmann agreed very well with those as calcu-
lated from the formulae of Abraham. I shall discuss
■t
these results later on in this paper.
ELECTRICAL EXPERI'^ENTS .
The electrical 'nethod, whenever prtictical, is
in some respects to be preferred to the photographic,
in as nuch as many readings can be •'ade and confirT.ed
in the same ti'^e tnat it takes to 'nake -i sinf;le expos-
ure by the latter. In my first experir.ents a sensi-
tive Dolezalek electror^eter fras used, but it *as
found a I" Oft i r^'.no ss :1 bl e to shield this and the
connecting v/ires fron electrostatic influences
sufficiently to obtain accurate readings. It was
therefore replaced by a sensitive t^old leaf elec-
troscope, which could be shielded very ea.':il3'' from
the effects of the hif^h potentials uced.
Expe r i ne nt I .
The fi'-st experi^nent was tried in air at atnos-
p.heric nressure, and the general ar'"ar<^e'':ent is shown
sketched in Fig. T. Two zinc plates 18 c's. long
were attached to insulating uprights •'*. and B, which
kept then in a vertical position about 9 r;"is. apart.
At the bottom of the plates was placed a block of lead
containing some radium bromide, covered ?/ith a thin
sheet cf .mica, vrr.ich allowed the 8 rays to pass through
^ Earth.
a I
Earth
A
B
Radium
VI 0,1-
'jtp.out 'luch abse?-T.t3on . The radium msls r,o arranged
that it' extrene edre -vay i'^ a vertical line twith
tne plate b, thus causing the rays to rraze the plite
b, but iJJowing the-! to fall fulD upon the plate a.
Over the top of the plates 'i^as sus^'ended the elec-
troscope, the details of >.vh3ch are shown in the
fjc'uro. The rod c containing tne fold leaf was
insulated from the case by -^eans of a bead of sul-
phur e. The syste-^ was charged to a potential of a
few hundred volts by means of the rod f, wnich could
be turned so as to touch the rod holding the gold
leaf. At other tJ-^-es it was con-iected to the case,
which was eartned. Tne bottom of the electroscope
had a thin aluminium windov»r throuph whJch the rays
could pass. The time which the cold leaf took to
fall thT-cufh a fixed distance on the cross hair of tne
telescope was taken as a measure of the ionization,
'^art of this ionization was due to tne 8 and secon-
dary ravs, and part to the Y radisAtions, and these
could be di sti nFuisned from one finother by absorption
tests. Of tie total about 60-^ was due to tne V" and
the remainder to 3 and cecondarv ra-"s.
If noiv 1-iie plate a i ;^ charped positively, then
the 3 ravs snould be deflected away from the plate
\.
b, unci a decrease of tne ionization observed. Triis
cecrease v/ould be a "teasure of the deflection of the
rays since no fresh rays could be bent in to take the
place of t-iose bent away . If the. rays travel -it dif-
ferent velocities, then tnis deflection ./i-uid be only
an averat^.e value, but by absorbing tne ones of lov^rer
veJocity one ourht tu be able to obtain tne deflection
of the hirn velocity rays very approximately.
The plates were "^aintarnad at a difierence of
potential by connecting them to a battery of small lead
accumulators, from v/hich a max.imum potential of 5000
volts could be obtained. In tne later experiments tne
number of colls rfas increased, so that a potential
difference of about 9000 volts was available.
"fhen a difference of potential of 3000 was applied
to the plates no appreciable decrease of the ioniza-
tion in the electroscope could be observed. The rays
were completely deflected by means of a mapnetic field,
but the value of HP., H being the strength of the mag-
netic field, and r tie raidus of curvature of tne
deflected rays, as calculated from this deflection was
over five times what it snould have i;een calculated
from t n e o r y .
The expression for the deflection of the ruy in
a uniform electrostatic field iej-
where 5 represents the uaflHct'on, d the distance trav-
elled by the ray in ^he field, h the distunce from
the top of the plates to the electroscope, e the
m
ratio of charge to nass of the electron, X the
strength of the electric field, and V the velocity
of the electron. In this experiment d = 18 c"-.s., h
11
= 7 ens. and K = 5 x 10 e • m . units.
If we as'^ume for V, and ^_, the extre"ie vt-ilues
m 10
found 1-jy Kaufrnann, viz., 2.8j x 10 cms. per sec,
7
f\nd .63 X 10 , the calculated value of the electro-
static deflection for th<t hifhest velocity rays
amounts to about 11 mm. v/hi(;h should have been
observed since the yidth of tne window of tne elec-
troscope vas only 15 mns.
This failure to observe the electrostatic de-
flection could not at first be satisfactorily ex-
plained, and it >vas only after a number of expe^ i^^ent s
by different methods were .r.ade that the true expla-
nation vas reached. I shall discuss this point later on
In this expsriaent , as in all otners made in air at
atmospheric pressure, .ve have no ver;,^ definite knov/1-
edhe of cue aniformity of the electric field bet.veen
the plates, and for this reason all the remaining
expsriments v/ere carried out in a vacuua. In tais
case the plates can be placed closer to?^,etaer, and a
auch higher potential applied oetveon the m v/ithcut
a dischar.tye taking place, tne field between tne plates
beinp; r-ractically uniform.
Experim ent II «
In fig. TI is sketched tne general arrange-
ment of tnis experiment. Two zinc plates, a and b, 6 c -g
long are insulated from one another, and kept a dis-
tance apart of 3.9 mms. by means of ebonite side pieces.
Beneath- tiie plates are fastened tiVo ebonite blocks,
forming betttecn tnem a slit of about 1 mm .vidth, tnrough
v/hich pass the rays from the radium contained in a cap-
sule placed below. The capsule is covered vith a thin
sheet of mica, and sealed so that no emanation from tne
radium can escape into the surrounding vessel.
Tnis apparatus was then enclosed in;:ide a glass
vessel k from Afhich tne air could be exhausted to a
high vacuum. The top of the vessel was closed by means
Elechoscope
/
C
D
'»*>n->"3-
'■'-^--'■'-1
Radium
F.gl-
of !i brass cap B, ^vaici .lud an opening cut tnrouph it,
and covered.vnth a very thin sheet of zir.c. T//o //ires
sealed t ii r o u g h the glass were connected to the two
plates, rind served to charj^e them to any desired po-
tential difference-
On top of the brass cap .vsre placed t'//o parallel
brass blocks, 3 cms. long and at a distance apart of
1 cm., v/hich supported the elnctroscope.
The rays fro"i the radium passed upward in a
diverging bean, through the thin saeet of zinc,a'"-d
into the electroscope. When the vessel was exh-tusted
to a high vacuum the plates maintained a potential
difference of 5000 volts without a discharre; in
this case we can assume that the electric field between
the plates is uniform. When this voltage, which corre-
sponds, to an electric field of about 17000 volts per cm,,
v/as applied there .vas a decrease of aoout 10-15 per cent
in tne ionization due to 3 rays. The rays >vhich reach
tae electroscope will consist in a larf;;e part of
those of r.edium and high velocities, sicca the low
velocity rays are nostly absorbed in the zinc plate.
If v/e calculate as before tne deflection which might be
looked for we find it to be about 3.2 m'-.s. for the
highest velocity rays, and since the distance betAfeen
10
the plates at t'ne electroscope is only 10 m-is^ tr.is
would a"iount to over 80^= conplete deflfctnon. Those rays
of lever velocity would be coapletely def ected. Haen
the apparatus was pl±ced in a nupnetic field the rays
could be completely deflected, though tne apparent
value of HR obtained whs over three ti"iss what it might
be looked for.
There v/ a s a slight possibility that 3 o m e of the
emanation might have escaped into the vessel, in which
case the true effect would be to sone extent masked by
the rays coning from the emanation. In order to prevent
this possibili-py anotner ar '-ange-ent was devised //here
the radium was kept entirely outside the vessel contain-
ing the charged plates.
Exper i me nt II I .
The experi lental arrnngement is shown in fig. IV.
A series of zinc plates, a , a , a , etc., IS in. number
were arranged parallel to one another, being insulated
and kept arart from one another by means of ebcntie side
strips. This a^range-^ent v/as enclosed int-ide a glass
vessel A, which had the wall? blown out in thin bulbs
immediately above and below the system of parallel plates^
Electroscope
Radium
FiqW
11
Alternate platPB .irere connected toretr.er and tc one
electrode, tne re"ia?.ninp plates bein?^ joined to the
other electrode. Tne lo-rta of the pl&tes .vas t c-,s.,
and the dist'^nce apart 1 mm.. The radium was placed
beneath the plates, and the electroscope above. By this
arrar.genent a :^.uch greater a'lount off? rays could get
into the electroscope than in the previous e xper 5 iisnt a ,
and also a large electric field cculd be prccuced
between the plates. A s all deflection of the rays
should be rBadily detected in tne electroscope. The
radium •ras placed about 4 c^^b. fro-n the botto'Ti of the
plates, so that the increase due to rayb being bent so
as tc ente'- the electroscope could not be great enough
to offset the dec-ease due to rays bent away and absorbed.
'Hhen s. difference of potential of 5000 volts ^vas ap-
plied to the pl:ttes no very appreciable deflection cculd
be observed. T::e magnetic deflection was also consid-
erably less than one r-iight expect froir. theory. The
total thickness of the glass walls wc<ld have about the
same absorbing power as tne sheet of zinc in the previous
experiment .
The results of these experiments frere at first very
perplexing, and no very satisfactory explanation could
12.
be reac.ied. One was loath tc concludo that tlie 3 rays
v/ere not deviable in an e Iri ct ro 3 1 at i c field, since Ktiuf-
rtjann in his ca-efully carried out experiTnents obtained
results which ivere in sucn extrenely good agreement
v/ith t:-iecry. Bftsitjes^the d i saf r e e-ient of the -na';netic
results .vith those obtained by other exper i'nen te r s ,
see".ed tc yhc'y th-^xt tr.e're .'/as ■:reBent Kcne p nenome na >vhi'jl
v/ere affecting both alike. ^Iothinf has bee>". said thus
far in these experiments concerning the effect of the
seccndar-'- >'a3''s ivhi ch are produced 7/hen the 3 and V rays
strike upon varr'ous substances. Tliit; effect vas froin the
first concidered, and vas thourht tc be toe c"iall to
a c c c u n t for the results obtained, b u t ^ a f t e r some f u r t '.1 e r
exper i r.ent s , tho conclusion 'vas reached tiiat it played
in all the experi-nents a lar^e role yxnd in sone cases
theprincipalone.
'^hen the R rays, cut down to a narrow pencil by
neans of slits, v/ere alloved to pass through an open-
in f in the bottom of the electroscope considerably
v/ider than the supposed 'vidth of the nencil, it was
found^on pi icing a narrow strip of lead over the opening
in different positions to cut off only a portion of the
p e ■: i n g at one t i ti e , t n a t the decrease in ionization
13
was the sa^^e for any part of t!ie opening. This sho'.vec"
cZearly that the. rays coirinr throuf;h the opeinr itere ?:ot
corfined to a narrow pencil as supposed but spread out
into a diffuse, unifor^ily distributed beam.
A photof^raphic plate placed over the opening siiowed
on d evelopr".ent a broad, unj f or-^ily dense iTnaf:e instead
of the sharp narrow one, which ought to be "lade by the
3 rays alcne. This broader. in- of the pencil of the
rays is unquestionably caused by the secondary radiation
produced by the pri-nary 8 and y rays. If this secondary
radiation was of the nature of yrays^then it wculd not be
deflectal)le in a nap-netic or rI e ctro st :it J c field, and
could con sequent I3; not affect the '^ ravs, but if on
the other hand it vas of the nature of the primary ? rays,
and therefore deflectable, great confusion would arise
fro^ the different radiations.
The secondary radiation produced by 8 -ind rays of
radium has been the subject of i nvps ti ^^at i on by ''cClel-
12 3
land , Eve , and Kucera ,who found that in general it
is of several kinds, both deflectable and n on- deflecta-
ble in a magnetic field, and of different penetrating
power. No definite knowledge was obtained of the velcc-
1. Phil. 'af. 4Ur]'id
2. Phil, "ag . tJhLC S'^o'^
3. .'Innalen der Physik. \*]oif
14
ities of the deflectable kind , and to this purpose the
author decided to -lake a complete i n vo s t irat n on of t'-ie
seco-idary rays. Before j'iving an account of this inves'
tip;ation I would like to describe t ato experiraants on
the 8 rays "lade by the pho to.c^r aphi c .-nethod.
PHOTOGRAPHIC E XPSRIHE MTS.
Ejep e r^3 m e_nt_ ^ . - The arr iinge'nent used in this is es-
1
sentially the sane as that used by Becquerel and is
shown sketched in fig. V, Two netal plates a and b
4.3 c Ts. in len,r;th \vere placed parallel to one another
in a vertrJcal position 6 m-n apart, and insulated by
'neans of ebonite. About 2 ens. belo/; the plates was
placed the radiu?n, ccntainsd in a n-irrow creva?se in a
block of lf"id. In the centre of the radium, -ind midv/ay
bet'sreen the plates was fastened a sheet of "^ica c, \v;-!ic'i
extended from the radiuT, up to the pho t o p;r aphi c plate e.
This plate v/as v/ranped in a sheet of black paper , and then
pl^jced Tn.-ide a lir^ht tight box containing a thin alumin-
ium wnndow throu,q:h v/hich the rays could pass without
"' u c h absorption. The distance between the top of t ii e
netal plates and the photofraphic plate "fas 2.4 c-is.
When the plates a and b are unchar'^ed there ourht to
be obtained on the p -lO togr afthi c plate an imaj-.e cros:;ed
1 . 1 c . c i t .
F.gV.
15
in the centre by a nurruv/ band. On the other hand when
th.e plates are charged to a hifj'.h potential if the rays
are deflected to one side, the pho t c graphi c plate s^iould
sho'.7 a broad shadow cast by the nica i-creen. In A and
B, f:.?:. 6, are shown two photofraphs obtained by this
method, A taken ".vithout an electric filed, ■■ ■' B v/ith a
potential differe>Tce between the plates of 3000 volts,
corresponding', to a field strength of 6000 volts per cm.
Both photof.raphs appear p'^ec.isely alike and she?/ no indi-
cation of any shadow on the plate. One side in both
photographs appears darker than the other; this is on
account of t^ie mica screen not beinj: plviced exactly in
the middle of the radium, thus throwing ncre raTrs to
one side than to the other. In C & D are shown two
ph(-, + o^raphs taken with the screen pl;'ced exactly in the
centre, C bein<^, v/ithout a field and D v/ith one; and both
sides of the imape now appear equally (len^em These ph.o-
to'-raphs were all taken with the radiun covered with a
thin sheet of nica, so that no emanation could escape
and affect tiie plate. Experiments viore also tried with
the radiun uncovered, and the sane results were obtained
as before. Radiii.m when dry and at ordinary t e^^perat ar e s
does not give off verv -nuch emanation.
n^M
16
Experiment II.- The arr anfj.e'-ent used in thic oxp eri.-r.ent ,
•xn.d illustrated in fig. 7, is essentially t.ie t;a-r,e as
1
that iiEod by Kaufinann in his work. T'./o brass plates, a
and b, pliced parallel to one another inside a brass box
A , \'iere i r. i; u 1 a t e d f r o n the sides by ''' e n n s of ebonite
bushinf.s c and d. Beneath the ;ilates was ''as termed a small
speck of radiura bro."iide, 'xnd above the plates a t'^ick
ebonite screen •■vhich had a hole 0.5 mm cut in it, through
which the rays could pass in a narrow pencil. At the
top of tiie brass box vas pi •iced a pho to.rrraphi c plate
wrapped in a sheet of black paper. The vhole apparatus
was enclot,ed inside a 2^1ass vessel which could be exhausted
to a high vacuum. /"- ^^ ^^^^^ ^^"^^3^
Exposures of one , two and four d a^r s v/ere made, but in
A
no case could any sharp image due to 3 rays be obtained
on thie plate, the only effect being a darkening of the
v/hole plate due to y and oedondary radiations. Hhen
the radium was left uncovered the darkening of the v/ ole
plate .'/as mucii greater, due to the emanation diffused
throii^ho'i t \ ^-'^ ■: ?;sl .
In another ar"angement a larr;er quantity of radium
conta.-^ned in a lead cap.sule v/as pi .cec' outside the brass
box, anr: ♦^.hp r.-\-'o allov'/ed to enter tnrough a small hole,
1. loc. cit.
/,v///////^//////j/j////f//^mA
Fig^zir.
and thence through an ebonite diaphrasra pLuced just
ber. eath the brass plates a and b.
This faflsd to give any better results than before;
and all further attempts to repeat Kauf-'ann'n exper^nent
•vere then a'bandoned. Proba^.ly irith better pho to." r aphi c
Ekill in restrainins the darkening effect due to the
rays I mij^ht have :-.ucceeded in obtaining a clear imacre
of ";he p rays. The radi uni U£-.ed in thi. experinent •.vas sor.ie
kindly loaned ine by Profestjor Rutnerford, and »vas as
strong, if not stronger-, t.-ian that used by Kauf-:iann in his
last experiments.
SE CONP ARY RADIATION.
The experimental ar^'angement ui^ed to study the
-ecrndary radiation is shown diagr amat i cally in fig S.
\bout 300 ;nilligranmes cf radium bror^.ide of about 30000
activity 7/ as sealed in u very ti-in glass tube of 1.5 :;i!n
dianetsr and about 5 cts. long. This tube v/as enclosed
inside a lead box a, .-/hich had a ilit 5 c^is. long and
1 '^:-i wide cut in one side, tnu -.vails of the lead box being
of such a thickness that no ,3 rays could pass through.
This box .'/as ce.'^iented to tv/o thick ebonite blocks b and c
in such a nan-ner, that one edge of the i^lit coincided
D
^ Electroscope ,
FigMl.
with a face of one of the blocks as indicated in the
figure. ^ d^verrent bean of '^ - -- - t'-u^, f=-..r-ed fro-n
the clit.ar.cl vas limited in svidth by the .valis cf t.-.e
ebonite blocks. This apparatus v;as placed -O" thfcit the
rays travelled in a horizontal direction.
Directly above and purallt'l to the ebonite blocks
v/as supported a thick screen A ,con str acted of cardboard
and filled v/ith -ercury, and iiuving in tne centre a '^/ide
rectangular opening throuGh ^/nich the radrut^ons could
pass into a gold leaf slectroGCope placed above. The
substances whi en were to be stL^died as sources of secon-
dary radiation were pi iced beneath this opening in such a
position that the f3 rays fell full upon them. The tnick
nercury screen absorbed a large part of tne radiation;
and althougn it introduced a snail a.ount of secondary
radiation, this incretiee was r^ore
than offset by the
decrease of the rays. The ionizat:on in the electroscope
due to the secondary radiation fron: below would under these
conditions forn a -luch lar-er prop.-rtion of the .v-ole ,
tnan .v::en the full rays were present.
iVhen tne radiator d is not present , the ionization
in the electroscope is due to V rays and to secondary rays
K
produced by then.
Tnen the radiator is placed in position this ionization
19
will be increased by that due to the secondary rays fron
below. i-iy fixing rad'ators of different rnaterials in
tne pcaition d, and pl^-cin" screens of varying absorbing
power under the opening 3, a knov.'ledge of the relative
penetratiing pov/er of the. secondary radiations from tne
different radiators can be obtarned. "hen a t.aick screen
is placed so as to cut off tne 1 rays, then the increase of
ionization vjll be due to tne secondary rays caused by
thefrays striking tie radiator d.
Radiator - ^ir and -urruundinf Objects.
Total Rate
of Leak.
3.C1
6 .0^
2 .9-i
Z .85
,' .70
3 .28
3.29
Hate of Leak
due to Secon
dary Rays.
1.01
0.4-2
0.34
.25
.10
.68
.69
Absorb-
ing 'Ma-
terial.
j -t s h e e t s
I paper
; 8" ••
16 "
3-1: "
5 "
liUnab-
sorbed
Ray s .
100
41
34
25
10
Remarks .
'lo rad : ator at d .
Paper placed on
top of "er-
c u r y screen.
Paper placed below
screen placed over
3 rays.
20
Radiator - Zinc.
Total rate
pf leak.
10.20
8.85
7 .76
6.21
4 .7-t
■i . 09
3 .53
3 .28
Rate of leak
due to r, econ-
dary Rays . t e r .1 a 1_^
7.10
5.7 5
i . 66
3 .11
1.C4
.99
.4-3
.18
'absorb-
ing "a-
Paper
2 "
4 "
8 "
16 "
24 ■'
40 "
60
^Unab-"
sor bed
'"-■ ays.
100
81
66
44
23
14
6
2.5
Re •^. ark s .
In this set
of readinp, s
the absorbing
layers of paper
.vere placed
directly over
r ad i at or d .
R ad i at or - Zinc .
5.00
1,90
It
100
rays partly
4.51
1.41
4
II
74
screened by 24
4.18
1.08
8
II
57
sheets of
3.80
.70
16
II
37
paper
Radiator -
Zinc .
3 .82
0.72
II
100
P rays screened
3 .59
.49 20
II
68
by piece of
3 .42
.32
36
11
44
zinc .
3.31
.21
86
It
29
\
21
Total Patej
of leak .
3.82
3.30
3 .2 k
'•or re c ted I Absorb.! ng
Leak . L ' 'i- "t o ria 1_.
"6
3.10
Radiator 4- sheets of
4.9 6
4-. 18
3 . 69
3 .34
3.21
.20
.14
.00
Paper.
1.86
1.08
.59
.24
• 11
i^adiator 8 sheets of
5.29
Paper
2.19
Padiator 16 sheets oi
5.81
paper
2.71
Padaator 28 sheets o
6.4 5
4.8
4.13
3 .57
3 .29
3 .23
P 'ip e r
3.35
1.70
1 ,03
.47
.19
.13
Radi-itor ^^lass
7.52
4.8 6
Radiator Load
11.36
6.85
4.4 2
1.76
3 .20
3 .75
ICO
10
32
4
8
16
32
82
8
8
^Ilnabsorbed
Rays .
100
100
70
100
58
31
13
6
100
100
100
51
31
14
6
4
100
39
100
45
Remarks ♦ _ __
Zinc screen over
3 r ay s .
2 7inc screens over
^ r ay s .
2 Zinc Gcreens over
3 rays.
radiator removed.
Raddator Copper,
Total Pate Corrected
o^f L e aj: . I lieak .
8.78" '" |~5.68
^ad 2 at orlron .
8 .69
5.3 t
5.59
2.24
A b s o r b i n r,
■'aterial .
8
/' U n a ij s r b e d
R av s .
100
100
iO
j-rom a study cf these results .76 see at once that, s -;b-
stances wheii t; truck bv 3 rays eive out seconcary rays,
■.vh3 ch differ anongst thenselves bcfn in quantity and
penetrating po.ver.
In o^'der to produce the maximum a lount of secondary
rays the sibstancr, has to be of sufficient thickness to
completely absorb all the ? rays which fall upon it. The
amount of secondary radiation due to theVrays is only
a very small quantity of tne total, provided the thickness
of the radiator is only just sufficient to absorb all the
P rays. 'Hhen only the ^.ost penetrating, rays are allo^ved
to fall upon the radiator, the secondary radiations pro-
sryu a/w. (KA^c\^A^c^
duced ire of u .ic-. -.ore penetrating nature, than -vhen all
/\
the '^ rays are present. This is partly due to the fact
that tne secondary rays due to tne hij^h velocity rays
are -^ore penetrating, and also partly to the presence of a
snail a-i.ount of se c end ary v' ray s cau:ed by the primary rays.
\
23
The denser a substance .is, the [^ r e a t e r -Mill be the
a-rount of the secondarjr rays produced, -ind "he p;reater
their penetrating power. Of tnoae substances exa-'ined,
lead proved to be the most efficient, and also gave rajrs
of the Tiost cenetratdng power, while paper .'/as the least.
For the sake of comparnson I give in table II tne results
of an absorption test on t -le primary S rays, (v/hich of
course contained a certain a^iount of secondary rays),
together v/ith the results obtained for the secondary ra-
diation frcm zinc and paper.
TABLE II.
Layers ~i
Prinary
Secondary
Secondary
of Paper.
_
B r ay s
ibb'
f romZinc
"ibb'"'
from Paper -
ibb
2
82
81
4
70
66
51
8
52
44
31
16
33
23
14
24
22
14
32
15
6
4-}-
9 .7
5.8
6
5 .0
2.5
1
24
The penetratintT pov/er of the secc?Tdary rays fro"! zinc
does not differ very ^.reatly from that of the pranarj'
fl rays, especially for the f?rst few sheets uf paper.
The penetrating pov/er of the secondary rays frorn lead
is only slightly greater than th;it cf the rays from zinc.
In fig. 9 are drawn several absorption curves, both
for the primary 3 rays, and for the KecondarTr ra3/s.
.'V represents that for the P primary rays, li that for tne
secondary rays from zirc, C that for the second ar3'' fro"!
paper, I) t^ie secncary from zinc when the pri-^ary R ravs
have pastsed through 28 sheets of paper, anc": E that for
the secondary from zinc when the primary S rays have
passed through one shee' of zinc '''t is seen at once
from thfcse curves that the radiiitions are not homo-
geneous, but travel at differe-^t velocities.
It is '.veil known that a purt of tne secondary rays
are deviable in a magnetic field, and in the same di-
rection as the 3 ra^'s would be. The magnetic deflections
of the secondary rays from lead, and from paper, were
tried h" t'-^e method illustrated in fig. 10. Tiie primary
B rays from the radium enclosed in the lead box a strik-
ing the radiator b, produced secondary rays which passed
up into the el r c t ro scorje . Screens of lead c and d were
n^.iK
25
so placed that the secondary rays frcn the radiator could
juct touch the upper eAr'.f> E of the Tiercury e ere en. Kxper-
iments wore perforr^ied vvit:'. both lead and paper as radia-
tors, and in both cases it h'xs found, that the secondary
rays produced .v/ere def]pcted tov»'ai'ds the ri.'^ht, v/iien a
mannetic field was applied in such a direct?or. , that tne
lines of fcce can be !-epresented ag goi nr into tne plnne
of the paper. The results obtained in t.nic experi-nent are
expressed in tne f o 1 1 o v : n " t a i) 1 e (III).
T\BLE III.
(Field' "s t7e ng t h
1 Und e f 1 s c t e_d_ J' ay s]
lead
p ap e r
ion
100
66
58
36
6-i
21
28
11
li
8
10
90
180
270
380
450
Thsfje results show that the Keccndary radiations from
both "letals and insulators consist for the T.ost part
of negatively char?;ed particles. Those from insulators
are -^vre easily deflected than those from the more dense
metals, but the difference is net very pre at. The pro-
portion of non-deviable rays in the secondary rays fron?
paper is gr&atar tiian in that from zinc.
TERTI^r.Y MYSj^
When the secondary radiation strikes upon objects
there is produced a third type of radiation called the
tertiary rays. Thir. radiation has been studied in the
E/ecfroscope
Merc api/
A /
Screen.
A.
\y^
T^ L -' SP
Radium
F,g.T^
ze
present investigation for -sl few substances. l';- o -ethod
e-'iployod .vas precisel^^ the sane as that ured for the
seconcary rays, with the exception that thR s-^al] ?^lass
tube containing the radrura, instead of beinr placed in
front cf the opening in the lead box, was "lOved to one
side so that no (3 rays could emerge frorr the opening.
Under these conditions the f^ rays will strike the lead
walls of the lead box and -reduce secondary rays which
will travel cut tlirnugh the opening.
If :re nlace a radiator in the sane position as
before, tnere will be produced i" it tertiary rays,
which can pass up to the electroscope, anc cause an
increase of ionization. Tne penetrating power of these
rays was studied for lead, zinc, and copper radiators,
and the results are sxpressed in the follov/ing table.
TABLE IV.
Layers of
Tertiary ^
Paper
Lead
Zinc
100
100
4
59 46
8
35
25
16
17
10
24
6
,i
27
A.S in the case of the secondary rays, lead proves to bo
the most efficient radiator, and al&o gives out the
greatest penetrating^ rays. The penetratinp po.ver of
the tertdarjr rays is considerably lesf> than that of
the secondary, that for lead in the former beiir only
slightly/ greater than. for paper in th« latter.
To find out tvhether th.e tertiary rays were ceviable
in a mapnetic field, the arranjenent shown d iagr amat i cally
in fig. 11, .'/as made use of. The rnercury screen A .vas
placed so that one end -vas directly over the ends of the
ebonite blocks a and b. The electroscope B was placed
near to the edrie of the screen A, and had a v/indow c
covered with tb.in foil cut in one side, so that any rays
might enter without absorption.
"I'aen the screen c was placed in position the tertiary
rays produced could travel upv/ard and en«ter the electro-
scope. '.:0W by placinr tv/o screens, d and e, in the posi-
tions indicated in thf^ figure, the rays could be pre-
vented from entering the electroscope, but a portion
would still be able to travel up'vard. '.Vhen a magnetic
field was applied in such a direction that th.e lines
of force v/ere perepndicular, and going into the plane
of the paper, the ionization in the electroscope v/as
not appreciably altered, but when applied in the reverse
Radi'i
(SS
SSf
Fis-II
28
directicn there vas observed an increase in ioniza'ion.
"oreover, when a screen was placed so as to cut off
entirely the tertJary rudiatitn fro^i LpIoat, thi- effect
was still observed but to a -^luch less de ree. These two
^e^■ults indicated that thre was or e c en t ^ be si d e r: the
tertiary radiation from below-a radiation which seemed
to come from the -lercury near the /nndow of the electro-
SCO e. This latter radiation is uidoubedly caused by
the y rays when they emerre from the surface of the mer-
cury, and is the same as that observed u- Kve nher\ tne )r
rave were pirtly absorbed by a thick block- of lead. i3y
placinp sheets of paper in front of tne window, it /as
fo'ind that the penet ating pcver of this radiation was
much less than that of the tertiary; t sheets of paper
cut it down to 30^ , and It: sneets completely absorbed it.
It >'/as also obi-evred that .viien the mapnetic field vas
increased slowly from zero, a certain .-trength was
reached at vhich the increase of ionization in the elec-
troscope due to the tettiary rays from below just became
noticeable. 'hen this hap'^e -.ed ^t he ravr from below were
bent so that the extreme eC^re of tne beam just entered
the windoiv. ''.noving three points on the circular path
of the ra-i's the radius of curvature could at once be
1 . loc . c i t .
29
calculated, and .vas found to be about 17 cie. Vhe
strength of the r^aj^netic field .vas about 90 C . C- . o . lines,
so that ve pet for the tertiary rays a value of H_fv = 1330
approximately. These results shov^ that the tertiary rays
are deviable in a marnetnc field in the same direction us
the R anc secondary radiat:ion. They :iust therefore be
ner.'^-ti ve ly charged particles travelling at apeede only
a little lesii than the slo.vest p particles.
'.".'hen the screen f is pl'iced in the position inciicated
in the firfure, it .vas found that tae ionization in trie
electroscope v/as increased. This ■vas found to be partly
cue to the seccndar-"' rays produced v.-.en they rays strike
the screen, and also partly caused by the tertiary
rays froTi beloiv striking the screen, and producing a
fourtn type of rays.
That this latter ■'■adiation is present, can at once
be shown by placing in front of the blocks a and b a
screen, sufficient to cut off all the secondary rays,
in which case the fourth rays are absent. This process
'.vould probabljr go on for a large number of radiators,
but after the third the radiation is too feeble to be
measuT-ed accurately. Each type of radiation is of less
penetrating power than the one v/hich produced it.
30
Velocity and ratio e for 'the primary 3 rays.
It has just been shown th-tt the secondary 3 rays
have nearly the sane penetrating pover ■xs the primary,
and therefore, assjuning for the r^resent that they possess
the same charge and mass, they "lUst travel at velccities
only sl^phtly less. After passing through a certain
thickness of absorbdnp; iniaterial, the lever limit for both
types of rays ^ould be the sarne , and consequently the
forv/ard edf^e of a deflected beam would consist of both
primary and secondary rays of the same velocity, which
■,7ere just able to pass through the absorbing layer and
cause ionization.
As the thickness of the absorbinp layer is increased
the less penetratdnp; radiation would be -nore quickly
absorbed, and at vsrv thick layers, the ed?,e of the
deflected bean ,7ould consist almost entirely of the ".ore
penetrating radiat'cn.
If then by an^ method '.ve ca ■ find the forward edre
of the deflected beam in both mapnetic and e 1 e c t ro s t ;.ti c
fa elds, ve can at once calculate the velocity of those
ra^'s, both primary and secondary, 7/hich are just able to
Jl
c aus eJ-cni Zfit i on after p. --is sin? thrcup-'o the absorbinp-
layeTt The folHovinp method .vas t.en devised fcr this
purpose anci proved successftl. It is sno.vn illusr.trated
in fdg. 12.
Two zinc plates, a and b, were supported in a
vertica] direction, a^.d insulated fro.-n. one another by -neans
of the ebonite blocks c and d. At the bottcn of these
blocks //as fastened tne lead box containing the rlass
tube of radium, tind f urni phed with a slit 1 ntiin in vidth
thrcuph vhich the rays err.erred. A.t a short distance
below the zinc plates there vas pi teed a lead diaphragm
with another slit also 1 mm in .vidth, .vaich for-^ed with
the opening below a narrow pencil of rays, ./aich travellee
upward uetvveen the zinc plates. The distance between the
two slits was such that the rays from the radiua just
touched the top edjres of the plates. This apparatus
was placed inside a pilass vesr-el A, which was clcsed
at the top and iiottom with brass plates B and U . Ihroucrh
the top plate B there .'/as cut a rectangular ope: in g
1.5 cms in iriCth, and then covered .vith as thin a sheet
of mica as would stand the pressure when the vessel .vas
exhausted. The thickness of this mica window was about
equivalent in absorbing power to 6 sheets of ordinary
writing paper. The bottom plate C was pierced with three
Electroscope-
F'lg^IK
holes, into which plass tubes were fitted, two of then
f and g containing the connecting ffires K and 1., and the
third H serving to connect the vessel to the pump.
The whole appn-ratu<j v/as rendered aj r tipht by tne
aid of sealinp --vax; afid it v/as founci^taut ;iixn an occa-
snonil stroke of the pump the pressure could be -lain-
tajned lc,7 enough to withstand a potential difference of
10000 volts witnout a discaarge takin;^ place.
'6v c o n n H c t i n g the .vires K a n r ' L to tne poles of the
battery of lead accumulators, the zinc plates could be
charged up to any desired difference of potential. In
the meantime the nunber of the colls had been increased so
that a naximun potential of 90C0 volts could be obtained.
On top of the brass plate B v/as placed the mercury
screen E, described in the previous pa<^es, which served
to cut down in a large measure tne y rays. Tne opening
in this screen was partly covered at the top by a brass
plate E so placed that no P primary rays could enter the
electroscope. The ionization in the electroscope would then
be caused by V" , and secondary' ra-"s.
The vessel was placed bet'/een the poles of a larre
electromagnet, of such a size that the rays during their
entire course la" in a uniforn field.
■6'6
When the primary rays strike the sides of tne slits
in the lead d i aphrafrn-is , neconclary rayc -ire produced wiiich*
travel ur>\/ard -.rith the prinar3' to the e 1 p c tro sco e . T.he
pencil of rays at the electroscope is then -nade up of both
primary and secondary radiation, but will be prevented from
enterinF<ahe brass screen E. This nencil of ra^s '*ill,
ho.'/ever, be confused by the presence of other secondary and
tertiary rays coming from various points. This can be illus-
trated by some photographs taken at different points alonp,
the path of the rays. One taken 1 cm from the upper slit
shows a narrow fuzzy imape, vhereas at 1 mm from the top of
the zinc plates the image is sharp and the full •Yidth of the
opening. \t the top of tne plate B the ircafe had again
become indistinct, while at the top of the mercury screen
it was of the full width of the opening and of uniform
density.
If the opening at the electroscope 'vas arranged
symmetr - cal ] y with regard to the zinc plates, there was
obfLeT-ved a decnase of ionization in the electroscope
when the magnetic, or electrostatic fields were applied
in either direction. The decrease in the case of the
electrostatic field was about 25 per cent, with a strength
of field of liOOO volts per centimeter. Thir, snows
clearly that the rays are deflectable in an electric
field, though the effect observed was onl'"' about one
quarter as Kuch as n;ight be expected.
■Vhen , hovever, the ar ranger.ent was the sa-e aa that
described in fip, . 12, altopether different results vere
obtanned. If tae plate a were charfred nefr,at a vely , the
ionization increased gradually to a naxinum wit:-, increase
of field, remained constant for a period, and tnen de-
creased. If, on the otuer hand, the plate a is positive
the ionization decreased at once. Tne reason for this
can be explained in the follo^inr manner.
If the plate a is charged negat i vely -the primary and
secondary p rays will he deflected to the right^and enter
the electroscope -thus causing an increase of ionization.
This increase .'/ill continue until the rays have just
reached the o p i j o s i t e side of the opening into the
electroscope. ''!her\ this is the case //e can assume that
those rays which after passing tnrough the; :» - c a o r b i r. g — lay o t
can just affect ionization, are bent the most, since
all rays of lower velocity, and hence penetrating yover
cannot get t arc a gh . \ '> \o - 1 '■.y fiV - o The same phenomena
exactly ,\rere observed \'tat^n a magnetic field was apolied.
Consequently whfin the ^.aximum point is reached for both
magnetic and el p ctr o s t at ic dfi f 1 cct ion , the value of the
velocity of the L-lo./est ray, which can just get through
the ajsorbing laj'er, can at once be calculated from tno
values of the electrostatic and magnetic fields. V/hen
^^ ^
35
different thickn eases of absorbing layers .'/ere placed
over t.ie openi^ig, the strength of fields necessary to
produce the .-r. aximun ionization .vaa found to incr^-ase
.vith increasing thickness of layer. T'r.ic is //hat we
would expect if the rays are of all different veloc-
ities, and cho 75,that v/hen the maxirau-n point is reaciied
the least penetrating;; rays for each successive layer
have all been bent through the sa 'stance.
For the sake of illustration sone of taese results
are expressed in the ac comp unyin<^. table (V).
Through
3 sheets paper
\i4 sheets paper
1.5 mms
. glass
r.PLVolH)
/ia+e o/ Ucck K
Pf6^vo(+6^
(ial€o±LenK
mi^Qiti)
(laieof^l^e^K.
ll'l- .Osec.
151.0 sec.
154 sec .
2700
13 4.4: "
3 400
l!r4.2 "
rOOO
tl5 "
3 000
• 13 2.4 "
3 6 00
143 .4"
5 6 00
14 3 .4 "
3 3 00
131.G "
3 9 00
142.8 "
58 00
148.6 "
3400
131.2 "
4100
141.5 "
6600
148.4 "
3 5 60
131.0 ••
43 00
141.5 "
68 00
143.6 "
38 00
13 0.8 "
C.O0O
141.4 ••
5 000
13 0.6 "
6000
13 0.6 "
tax. 3500 Volts -
4100 volts.
6000 volts.
In this method v/e do not kno// ho// far eacii ray is bent^since
we do not k n o v/ its first position, and therefore cannot
iib
Obtain values of jViV, and MV , ''Ut since the deflection is
s'liall compared to t'ne len.gth of tue path, we can assume wit^ioat.
any n;reat error, t:.it^ for a maximum, this is thw sa^ie for both
nac?netic and electric deflection. 3y comparing the e.ljctric
and magnetic forces on the particle v/e can a,t once esti';u.te
its velocitj^. If d is the distance travelled 03' the ra;' in
the uniform magnetic field H, 1 the lengtli of the zinc plates,
h the distance from top of plntes to olsctroscope, X the
strength of e 1 1- c t r i c field, the, v e 1 c i t j'- V is g i v e ^1 by,
V
Hd
j-
In this experiment, d = l?.*^' cris. 1 = 8.4 cis , h
the distance apart of the char.t^ed plates was 6.2
fore ,
J.J c
na
mms , there-
As an example take the results for an absorbing layer of
11
6 sheets of paper; here X. = 3500 = 5.7 x 10 C . C- . S . and
0.C2
K = 13.5 C.O.S. lines, and therefore ve obtain,
11 10
V = 0.5 63 X 5.7 X 10 = 2.37 x 10 cms / -ec.
13 .5
In order to obtain values for iS_, it is necessary to know
either Tm . or V\V, for the different ravs. Values of TnV
6 e e
were obtained by the folio./ JnH method.
37
The lead dox co 't -ii nine; tht; radian tube vas fastened to
the ebonite blocks as before, but tne di. stance between
these narro.ved to 2 n.'ns , and the upper lead disphragm
removed. This ar ''•i.np.enent allo.'/ed a broader pencil of rays
than in the previouL- experiment to pass upward to the
electroscope. Tv/o zinc plates ii cms in height ^rere fas-
tsned to t;ie ebonite blocks in a vertical directior: and
so arranged that one edge of the pencil of rays could just
touch the top of tne one of tne plates. The effect of this
upper plate ivas then to define siiarply a pencil of both pri-
marjf and secondary rays. This arrangement ^as then sur-
rounded bv a magnetic screen of soft* stieet iron v/hich ex-
tended up to the top of tne z«nc plates, and kept the rays
during theJr passage bet.veen the plates from tha action
of the Magnetic field. The glass vessel was dispensed v/itn
anci thfi pxneriinent perfornied at at:r.o spheric pressure, sheets
of paper equivalent in absorptive po.ver to the "ica v/indo.v
of the •previous experiment being interpcsuul in tne path of
the ra^'s. The opening at the. el ac tro scope , 1 cm in v/idth
was arranged as in t)ie other experiment so that no rays
of this pencil could enter ':he electroscope.
'Vnen a marnetic fielf' v/as :inolied in the proper di-
rection t:.e ionization in the « 1 i-ctro s cope gradually in-
;i8
creastid to a •naxi.'tura as the, strenptn of the fielc^ '.ras irr
c r e a 15 e d
1 n c e z
h? rays are only under the action of tns
nai'netac field during their passage fron the top of the zinc
plates to the electroscope v;e can as g u^ie t .lat ^ '.vhen the naxi-
mura po^nt is readied, tne leaat penetratinp rays have been
deflected over the distance represented by the '.'fidtn of tae
opening. By placinp, in the path of tne rays suitable thick-
nesses of absorbing 'TiaterDal the value of JJlVfor the sane
e
rays as were ob.served in the previouL^ experiment could be
ascertained.
These values 'vere calculated fron the f ol lo .'/in? for-
nula , " ( See J. J. Thomson's, Pischarf^e of Electricity
through Gases.,, pa r, e 92.
//hereO= the distance taroafh ./hich the rays are bent, and
H the value of tne "lapnetic field at d:^fforent porlnts along
the Tath of t n - ravs. The value of )Tiy obtained bj' t.'iis
^ " 3 ~e
;-ethod was 1.57 x 10 for no absorber, whereas, for 1.2 '^.m of
3
zinc it 7/as 5 x 10 ; for C siieets of paper corresponding to
2
the "^.ica v/indow tie value ".vas 1.87 x 10 .
In the forr.er experi"ient in the uniform field the value .-tf*^"^
of yyW- n (^ , and substituting the • ; ?»< > ■». « valuer of IfYW/^ .'/e ret
for the distance thro'irh vhich the rays .vers bent 1.04 cis.
Knov^ing this value of u all the remaining values can at once
be calculated, and the complete results are sacvn in the
39
follov/inr; table (VI).
T.\3LE VI.
6 of paper
10
14
18
22
3 "
X(nv^>)
1
11
5.7 xlO
6.1 "
6 .G "
7.2
7 .7
8 ..i
1 . 5rnm g^ ass) 9 . 6
13 .5
l-t.3
15.1
16.2
17 .0
17 .9
20.0
•2^
.3 coppef
0.4 zinc
.8 "
1.2 "
11.0
li.O
22.2
13
4.4 xlO
4.76
5.15
5.62
6.01
6 . kl
•1 . 49
8 .58
3
27
.2 0.0.9 2
1.57x10
1.87
1.98
2 .10
2 .25
2 .3 6
2.49
2.78
3 .09
3.7 9
4.6
5 .00
'^Cm^j,
Uc
2 . 2 ; e s -t . )
10
2-37x10
2.40
2.4 5
2 .49
2. 54
2. CO
2 .09
2.7 7
2.88
2 . 9 5(e E t i -
2.97 '
m
1.40
1.27 xlO
1.21
1.17
1.10
1.07
1.04
.96
.90
.76
.64 •
.59 *
i
-\s has alreadv been exnla? n ed ^tne x'alues f^iveti in the above
table .represent those for both the primary and secondarjf par-
tides 'Vh:' ch are j'.'nt able to penetrate thro'uph a certain
absorbing layer and still cause ionization; the highest
values of the velocity are tncce for tr.e !3 particles alone,
since nearly all tne secondarj' p'irt:cles are absorbed in about
• These estimated values were extrapolated fron the curve
showing the relation between the velocity and >nV.
40
35 sheets of paper. The values of the velocity for th.e par-
ticles thro'^fTh 0,8 and 1.2 mm of zinc could not be ob-
tained fo"* the reason th;t tne linit of the available) dif-
fsrenco of potential nad been reached. The valuer of ihWfor
e
th?se, ho '/ever, show that the apparent -nass of the parti-
cles ir increasing rapidly as the velocity of lif^ht is
approached. The In.ver liTiit of the ^ rays camot be stated
■vith anir certainty but it i c probably greater than 2.i x
10 "
10 cms per sec. T'ae *? rays in a]l the oxneri-ents had
fnrst of all to pass- thT'oagh the thin glass walls of
the tube, v/h:5 ch would probably absorb about as "luch as the
n i c a w i n d o Af .
The results f.iven in the above table apree vev' veil
1
with those obtained by ''auf-^ann in hi r later experi-
■"ents, usinp; an entirely different method, and shov beyond
doubt that the p particles from radium travel with speeds
approachin,^ that of ]ight, nn6 further that their apparent
mass does increase with the speed.
In Kaufr^.ann's experiment secondary radiations must have
been present in quantities quit© comparable with the prim-
ary, and consequently any point chosen on his curved pho-
tographic traces would represent not alone the primary par-
ticle, but -^Igo some secondary particles v/hich happened to
arrive at the sam.e point. If tr)e pri'iary and secondary
1 . lo c . cit .
il
particles differed i 'o velocity there -njrht p...«ibly be
sor^e ,^dff:caty in .st5-natin? tn. radii of curvature for
the different ravs. He h.. cmpared his results ^i th
those calculated fror. the theoretrca] formulae of Vora-
ham, and fines a very close agreement, thus giving con-
siderable evidence to tne view that the -.ss of the elec-
tron is entirely electrical in orison, and increa.es as
the speed apnroaches tp.at of liR^nt. The result of the
present in vost i ~at i on also see--,s tc l.ad to the sa:^.e
conclusion. 'Ve ^ust, no .ever . re-nenber that the confirna-
tion of tni. tneorv is only over a short ran^e w;>ere tne
increase of the aonarent -^ass is not very -reat, and
vhere any F,et of values niPht accidently agree very
closely v;ith the theoretical. It is not until the
speed has. for all exparimeatal purposes, practically
reached that of lip;ht that tnere is any ^reat increase
of tne apparent mass, and consequently any conf i rn^.at ion of
the theorv over a long range is entirely beyond exper-
i-.ental nieans. "'oreover the a.-sumptions -ade in the
deviation of tn.- theoretical forniulae, and also in the
for-nulae used to calculate tne velocities and ratio e^
from the experimental data, are many and some'Vhat grea^.
It is assumed that t ,« same laws of electric and magnetic
42
force apply equally well as very hiz'n speeds as at com-
paratively low, somethinfr of .vh: eh wo have no a priori
knowledge. ^ diver "ence of either, or both, of these laws
at h5(^h speeds could -iccount for the ax^ er i-nent al fxcts
observed .
Hovevor considering; al] sources of error the re-
sults of Kauf raann , together with those of the author
lend great weij^ht to ths view that the '"aso of the
electron ic electrical in origin, and increases with
the speed .
Velocity and ratio ^ for the secondary 3 rays
"7n
It has !)een assuned in the previous sections that
the velocity of the seccnOary rays was only slightly
less than the primary, and also that they carried the
sane chars-e and had approximately equal Trasses. That
these facts are true was proved by the follov;:ng set
of experiments.
The general arrangement of the experiments vas the
same as that used in the case of the primary rays, and
the same drawings and general description will suffice
here. Thn tube of radium, instead of being placed under-
neath the slit in the lead box, .vas shifted to one side
43
so that no primary raj's coulri emerge fron the box. The
lead box was fastened to the ebonite blocks so that the
slit .'/as near the rdpht hand block, and net in the ".iddle
as v/as the case vitn the priTiir;' rays. The upper lead
diaphragm :ra.B re'.cved and in its stead was fixed a strip
of lead , which covered one t.aird of tie opeiinF belA'een
the blocks. T!ie secondary rays on energinp: fro-n t.ie slit
in t \e lead box could strike full upon tne left hand
zinc plate ^but were prevented by this strip of lead fron
falling upon the right hand plate. The rest of the ar-
ranc^ement vas p'-ec-'sely the sa-ie as for tae primary rays,
the opening at the electroscope being arranged so that
no secondary rays from the lead box could enter.
The secondar''^ rays when they strike the left-hand
plate -,and also the edge of tlie lead .trip produce ter-
tiary rays which travel up vith the secondary rays. \
great part of those will be absorbed by the rr,ic.a window
but so-ne will get through and cause ionization in the
electroscope. Tnere U. also a s'nall ar^cunt of secondary
rays produced by the rays striking the various parts of
the apparatus; these, however, do not affect the results
since the- cone fror. all different parts of the vessel,
and are therefore deflected verv unequally.
•tl:
When thtt elactric or -agnetic fiolcls vere applied in
the sa-^e directions; as in the case of the primary ravs
the sane phenonena were observed; a fradual 3r,crease of
ionization to a maxinum //ith increasing fields, in one
dnrectfon, -xne i decrease at on
ce wit!! -Vl.e field an th
other. The sa'^e abgcrb5np laj'ers were used in this case
as in that of the prf^-nary rays. It vas fuund that no
increase of denization above 30 layers of paper could be
Ob erved, that thickness being sufficient to alr.ogt
completely absorb the secondary rays. These results
confirm the idea that tie secondary raye have a less
penetrating character than the primary, and travel at
lever cT.-'^.nds.
Tn order tc pstin^ate how far the ray£.= rere bent in
tne electr^-c and magnetic fields the same metliod .vas used
as for the primary rays. Al ;! that ./as necessary to do 'Has
to shift the tubo of rad-:an from beneath the slit in the
lead box tc one aide so that no primary rays could get
out. The soft .iron .ecreen v/as placed ,0 that the rays
■rere shielded from the influence of the field until they
had passed the top of th. zone plate. The distance from
the top of the zinc plate to the electroscope //as less
than in the case of t
he primary rays, but tie ffidth of
the opening in the electroscope the same. The value f >Tff
e
1-5
3
for r. c absorbing layer v/as fo and to be 1.56 x 10 ^ and
for an a b sorb ring layer of 6 sheets of paper v/as 1.85 x 10 .
These values a.re about the sa-ne as the c o t espc nci mg ones
found for the priniary radiation, and she* that the lo;/8r
limits :r ^'-.e latter case were those of the secondary
rays. The upper limit for the secondary rays by this
lethod .'/as also thp. gar-.e as by the otner method, viz.,
3 1 ay e r s of paper.
r u b s t i t 'u t : n f^ t h « s e values of JTW in the; results
e
obtained by the fi»"st method v/e obtain 1.29 ens as the
distance throur^h .yhTCh the ""ay ■ were deflected.
In thic; experiment, 1 = 8 .-i: cms, h = 5.5 cttib , and
d = 17,3 cris, :^r\6 therefore,
In the. ac coripanying table(VIT} are expressed the final
results obtained for the secondary rays.
TABLE VII.
\bsorbi ng
"' ^
mY \
1
V cirs/ sec
.^
La;;er .
X(:nax)
H (nax)
e '
m
3
+
11
-
1.56x10
10
7
C paper
6 .9x10
If,
1.85
2.35x10
1.27x10
14 "
7.5 "
17
1.97
2.10
2.50 ■'
1.21 "
2 2 "
8.5
18.5
2.16
1.16 "
3
9 . 6 ••
30.::
? .36 ;
2.58 "
1.03
46
It will bei seen at cnce fri.m tnis table that the values
of t'le velocity and ratio ^_ for the secondarv ravs are
m
in t'ood apreernent with those obtained for the primar'/.
"ince the exr>eri-^ents on the secondary radiut.ions fru'n
different substances, described earlier in this paper,
showed th-it they were ?.ll deflected in a magnetic field,
and since further that the penetrating power cf the rays
vfrere not ,-reatly different, we can conclude that both
metals and dielectrics J'ive off when struck by rays,
negatively char-eo p^trticles .fi^h velocities, ;\r\c ratio
( thp san.e as these for the less i^enetratinp S ravs.
m
■-.s to the origin of tnese secondary and tertiary
radi at i^'ons , twc possible explanations can be given. They
Tirht be the pri-:ary B p-irticles reflected back fro"i the
"lolecules of the bonbarded substance. The ether and
-luch more prob-i'De vie>v is that, the secondary radia-
tions are pruduced Jfhen ine pri-nary 8 ritys, striking the
r.toms of the substance, cause them to give off one cr
nore of the electrons of //hich they are co-nposed, v/ith
sufficient velocity to escape fr^m. the substance. The
longer the path, and the greater the '^.umber of atoms
which a particle encounters before it is absorbed,
the greater ffil!! be the a-^. unt of thu secondary radiation
produced. The nature and a'lount of the secondary radia-
17
tipne doMthe>iatiireoft'i'-atfc'". of
♦'. n production of
socontiar; tertiary rud ' is t .en . inilar in rna'r;
reelect 8 tr. fl action of P rays by the radio^ictive
i ul • '■ • • ■ '• * " ( " i - -IS f ttT as
*e '-.•oti ^ ro n ♦ n-' Hi.u •• for"ipr JV is tiue to external
agents.
It *jll pern.'ips Lie instructive * ' e f ly sun.iiar-
ize 'rxfiTo the main facts brourht out i- iv-
\08tif-it3cn.
The d5fficilty exp er n en ced at first in ootai -n
L'l '^ctrc a t 'it i c defl ct itn of * le "^ riyt c^m to be
secr.cary anc te'^tiary r ad J it i o n l. •
''I'.e-i til is disturbinp, ictjicn had been t itec by
f M r -,-<*-, f' Q •ion . ^ * . c i t y , -and
'■ A t i o ^ , ./ e r e o u t a i n e c c s e 1 y * i t h t h t s e
frund by ''aufmann usinr an entirely different "lothod.
_ (» 1" •*(>•( ^ H r II'"
a '^ e produced s .a- p r i r;
inz rate'"?al, <-re conpticed of
rZ r z t r ■'.".'{' 1 1 i :'
,or-.,r-- TH'"" +'fns, y/hich
s t,trike -pon an absorb"
ively charred pxrti-
t? :: nnnrl'- f? T •■ "! tc '-.it . <■ the
18
rays t ne-ise Ives , and hivinj^ the La"ne r-itio jS^ for corre-
spondinp velocities. The ;: e c one' ar ■' vays t he" ce 1 ve s ^v/hen
jncr'dent upon a sub stance «prc d u c e tertiary r ay 3 , vTh5ch are
of slightly lesf; penetratinp, c hrir i^c t er , 'ire deflectable
in a •naf.netic fi^ld in the sa-ie di'-ectnon us the secondary
ai(i pr.-:~i>.y « rays, and are probably of the same nature
as these.
The values of _^ obtained for both the ornmary -xnd
seconcar" n rays shoived, that t^ie mi^-'i ^r e n t "^'iss of tne
part:'cles increased as that speed api'roacnsd the velocity
of 1 i f h t .
Taking f o - <-»-Hnted that the asf;umptions involved in
t h ? c a 1 c u 1 a t :> o n s of the v e 1 (> c .i t y , "1 n d ^ , are correct,
nv
•ny •results together .vita those of "aufmann fur-iish evi-
dence in favour of the view that tne maci. of the el;=ctron
is entirely electrjcal in nature.
In conclusion I vish to express my tincere thanks to
^rofessor \?^!es for r.is kindly interest, valuable t.ug-
festJons, and for the facilDties -.V'lich he Placed at '^,y
comniand during the pronress of this i nve sti !-at i on .
Johns Hopkins 'University,
Balt-imore , "d .
''arch, 19 PC.
