JOURNALOF GEOPHYSICAL
RESEARCH
VOLUME 55, No. 3
MARCH 1950
The Shadow of the Earth's Core
B. GUTTENBERG•'
SeismologicalLaboratory
California Institute o• Technology
Pasadena,California
Abstract. There is very good agreement between the observed decreasein amplitudes of
longitudinal waves diffracted at the boundary of the earth's core and correspondingtheoretical
results. Especially, no rapid decreaseof amplitudes at the beginning of the shadow zone is to
be expected. Theory and observationsshow that the amplitudes of diffracted short-period P
waves decreasefaster with distance in the shadow zone than those of longer waves. At epicentral distances of over 110ø short-period diffracted P waves emerge gradually and their beginning can rarely be ascertained. Corresponding long-period waves arrive within the limits of
error at the time calculated on the assumptionof a straight-line travel-time curve.
Amplitudes of waves diffracted from a caustic decreaserapidly with distance; this usually
limits the range of their observationto roughly 10ø. On the contrary, amplitudes of P waves
diffracted at the core may be visible on records of a great earthquake at distancesof 103ø to
180ø. Observationsfor diffracted S waves cover a much smaller range of distances,since S
wavesappearin a portion of seismograms
whichis disturbedby earlier motion. Otherwise,observations indicate
similar behavior
for diffracted
Introduction. In his first investigation of
the earth's core, the author [Gutenberg, 1914,
p. 192; 1925b, p. 23] explainedthe observed
longitudinal waves at epicentral distancesbeyond 102ø as waves diffracted around the
earth's core.Althoughthis hypothesishas been
widely accepted,it has been doubtedby Lehmann [1953, p. 303; 1958], who suggests
that
the small amplitudesof P beyond about 104ø
may result from a decreasein wave velocity at
depthsapproachingthe core.
Records of the 'diffracted
P and S waves.
had a magnitudeof about m = 73A, correspondingto about M = 8•. It was studiedby
Macelwane[1930, with reproductionof seismograms] and by Krumbach [1934] who reproduced seismogramsseparately [Krumbach,
1931]. The diffractedP is clearly recordedto a
distanceof 162ø (Paris) and doubtfully[Macelwane,1930,p. 214] at 167ø (Oxford).
Fundamental data. According to Jej•reys
[1959, p. 98] the mean radius of the core is
3473-----3km. This value is based partly on
P waves' at dis-
travel times of longitudinal and of transverse
tances over 140ø are rare, since they require wavesreflectedat the core boundary.The flatlarge earthquakeswith epicentersin regions tening of the core is about 1/460 [Gutenberg,
which have distances of 120 ø to at least 160 ø
1959a, p. 162], so that the differencebetween
from many well-equippedstations.Earthquake the equatorialand the polar radii of the coreis
epicenterswhich fulfill these requirementsare probably between 7 and 8 km. Accordingto
in the western East Indies with the recording Jef/reys [1939a, p. 511], longitudinalwavesarstations in North America, and in the New riving at an epicentral distance of 0 -- 90ø
Zealand area with the recording stations in have their deepestpoint at a distancefrom the
Europe. The most extendedrange of usefuldata earth's center of about 3660 km, where their
seemsto have been furnishedby the shockof velocityis 13.61km/sec. Rays arriving at 0 =
June 26, 1924, at 01:37:34 UT. This originated 96ø have their deepestpoint at a distanceof
at 56øS 157$•øE (near Macquarie Island) and about 3545 km from the center; the velocity
there is 13.64 km/sec. This velocity remains
x Contribution No. 957, Division of the Geoconstant, according to Jeffreys, down to the
logical Sciences, California Institute of Techcore boundary.The longitudinalwavesgrazing
nology.
? Deceased January 25, 1960.
the core arrive
1013
at a distance of about
102 ø.
1014
B. GUTENBERG
Gutenberg
andRichter[1939,p. 134] founda
velocity of 13.8 km/sec at a distanceof 3570
km from the center and from there a small de-
no need for suchan assumptionif there is no
clearly definedshadow,and the result of the
present investigationis that there is not.' On
creaseto 13.7 km/sec at the core boundary. the otherhand,Lehmann[1958, 1959] pointed
They calculated
that.in shocksstartingat the out that at distancesup to 108ø short-periodP
surfacethe longitudinal
wavesgrazingthe core waveshave beenrecordedjust as well as longarrive at an epicentraldistanceof about 103ø.
periodwaves.In addition,she pointedout that
Transversewaves arriving at distancesof the conditions for the transmission of P and S
91.6ø, 96.2ø, and 103ø have, accordingto Jel•- wavesin the portionof the mantle immediately
]reys [1939a,p. 513], their deepestpointsat above the core differ, sincebeyondthe critical
distancesof 3667, 3549, and 3486 km, respec- distanceof about 105ø the amplitudesof S are
tively, from the earth's center, with corre- frequentlyconsiderable.
spondingvelocitiesof 7.26, 7.31, and 7.30 km/
From the precedingresults it appears that
sec. Gutenbergand Richter [1939, p. 134] the explanationof the P and S waves observed
founda decrease
in velocityfrom 7.3 km/secto beyondabout 103ø as diffractedwavesdepends
7.25 km/sec in the 125-km-thickportion of the on answersto the questionsof how amplitudes
mantle immediately above the core. All these of such waves decreasetheoretically in the
rates of decreasein velocity are definitely shadowzoneand whetherwe have to expecta
smallerthan the critical rate of decreasegiven 'sharpedge'of the shadowzone.
Theory and model experiments concerning
by dV/dr -- V/r [see,e.g.,Gutenberg,1959a,
waves diffracted at the core boundary. Waves
diffracted at the boundary of a cylinder were
investigatedby Friedlander [1954]. He found
that. a diffractedplane unit pulseshouldfiatten into a wave with its maximum moving more
slowly than the beginningof the disturbance,
as the distance from the beginning of the
ably smallerrate than in the shadowzonepro- shadowzone increases.Jef]reys [1959, p. 90]
ducedby the low-velocitylayer of the asthenos- pointed out that this theoretical delay of the
phere [see,e.g., Gutenberg,1959a,pp. 76-89]. maximum 'appears to increase more rapidly
Combinedwith the nearly straight travel-time than linearly with the distancetraveled in the
curves,this indicatesnearly constantvelocities shadow.'
in the portion of the mantle near the core;
$cholte [1956] developedsolutionsfor somethere is no indicationof an appreciablede- what simplifiedequationsfor elastic waves in
crease in the velocities of P or S.
the earth. However, he consideredthe curvaLehmann's[1953, p. 293] main objectionto ture of the layers. I-Ie applied his solutionsto
the hypothesisthat the P and S wavesobserved the shadow zone [Scholte, 1956, p. 32] and
beyond epicentral distancesof about 103ø are found that the amplitudesof longitudinalwaves
diffractedat the coreboundaryis that P 'does with a period of 8 secshoulddecreaseto about
not seemto fall off suddenlyat this or a neigh- one-half at a distanceof 8ø beyondthe boundboring distance,nor doesits appearanceun- ary of the shadow zone. He concluded that.
dergoa suddenchange.'Moreover,shereported 'this rather slow decreaseexplainsthe fact that
that during her investigationof recordsof a P waves have been observed far inside the
shadow zone.'
Chilean earthquake 'serious doubt arose respecting the existenceof a shadow having a
Rykunov [1957] investigatedthe problemby
sharp edge.' She investigateddata for many using ultrasonic modeling. The similarity beshocksand concluded[Lehmann,1953,p. 302]:
tween his model and the earth is good except
'The fact that there are rarely short-period for the periodswhich, in the model, are about
waves in P at great distancesmay be in favor
ten times the equivalent of the periods observed in the diffracted P. I-Ils result shows no
of the assumptionthat it is diffracted,but it
cannot be taken to prove it. Actually, there is 'sharp edge' of the shadow. The decreasein
p. 25] for formation of a complete shadow.
Near the core boundary,this critical rate is
about 0.4 km/secper 100 km for longitudinal
wavesand slightlyover 0.2 km/secper 100 km
for transversewaves.On the otherhand,at a•
epicentraldistanceof about90ø the amplitudes
of P and S beginto decrease,
but at a notice-
SHADOW
OF THE
amplitudesis faster for short than for long
waves and dependson the rigidity in the core
near its boundary.
Knopo• and others[1959] includedthe problem of diffraction of elastic waves by the core
of the earth in an investigationof scatteringof
seismicwaves. They found that similar to the
diffractionat the boundaryof a cylinder,mentioned above, a wave diffracted by a sphere
getsflatter, and that its maximum is delayedas
the wave arrives at increasingdistancesin the
shadowzone [Knopo• and others,1959,Fig. 9,
p. 124]. Slightly over 30ø within the shadow
zone the amplitudesof waveshaving a period
of about 2• sec have decreasedto about onetenth, and they have a minimum of about 0.03
about 50ø from the beginning of the shadow
zone. In their calculations,Knopoff and his
collaboratorssuppose that the shadow zone
begins at an epicentral distanceof 120ø. Near
the anticenter
of the disturbance
EARTH'S
CORE
1015
quake of April 16, 1957, 04 04 04, in the East
Indies, adoptedepicenterat 4.5øS,107.5øE,focal depth 600 km. Recordsof this shockshow
unusuallyclear P and S waves (Figure la to ld,
lk), as well as PKP and SKS waves [Gutenberg, 1959b]. The author is grateful to many
stations for the loan of records of this shock or
for copies.All data used in the following discussion are
based on the
author's
measure-
ments.
In general,it has beenfound by investigators
of diffracted P waves that at epicentral distances of more than 103ø they frequently arrive later than the time calculated on the basis
which corre-
spondsto 60ø within the shadowzone,according to their assumption,the amplitudes have
increasedagain to about one-sixthof their original value as an effect of the concentration
of
energy near the anticenter. Thus it followsthat
in great earthquakesthe diffracted waves could
well be observed as far as the anticenter.
The results discussedin this section agree
very well. Theoretically as well as from model
experiments it follows that the amplitudes of
the diffracted waves should decreaserelatively
slowly from the beginningof the shadowzone
where no well-marked discontinuity should be
observed.Only a gradual decreasein amplitudes shouldbe found. Waves having small periods should show a greater decreaseof amplitude with distance than those with large periods. Approaching the anticenter, which has
the propertiesof a focus, the amplitudesof the
diffracted waves should increase correspondingly. The smallest waves should be observed
roughly 10ø from the anticenter where waves
having periods of 2• sec should show amplitudes of the order of 1/100 of those at the beginning of the shadowzone.
Observations. Seismograms showing diffracted P waves have been reproducedfor the
shockof June 26, 1924, by Macelwane [1930]
and Krumbach [1931]. In addition to these,
original records have been used of the earth-
(•P
©:•28:4
e:1o973
•ig. 1. Portions oœ seismograms of sI•ock 0]3
April 16, 19,57,0• 04 0• U'T, A: 600 km, a]] on
the s•me time scale; (•) long-period and (•)
short-period vertical components of P at Tamanrasset; (c) short-period, (d) long-period vertical
and (e) EW componentsof P at Kew; (•) vertical component of P at Malaga; (g) and (h)
short-period vertical componentsof P, (g) at
Eureka, Nevada and (h) at Woody; (i) ultra•
long-period vertical component of P at Pasadena;
(k) and (l) NS componentsof S, (k) at Kew and
(l) at Toledo. The arrows in (g) to (i) indicate
the time
at which P has been calculated
from
Gutenberg and Richter [1936, p. 349] for h -600 km.
1016
B. GUTENBERG
P colcul.
o)
e=159-ø
minute
-J
Fig. 2. Beginn•gof se•mograms
writtenby verticalGalitzin•ismographs;(a) at Eskdalemuir,
June 26, 1917,05 49 40, ba•d on orig•al record; compare Gutenberg [1925a]; (b) at Paris, June 26,
19•, 01 3726, based on Macelwane [19•, Fig. 30]. Estima•d accuracy of adopted epicentersabout
1•ø, of origin times about 6 •c.
of travel-time curves extrapolated by straight
linesbeyondthe end of the direct wavesat an
error at the time calculated on the basis of the
epicentraldistanceof about103ø.Jef•reys[1959,
p. 89] and Lehmann [1953] pointedout that
the beginningof diffractedwavesis frequently
doubtfuland may be too smallto be observ-
adopted elements for the shock and of the extrapolated travel-time curve. However, the beginnings of waves having periods of less than
about 3 sec (Fig. lg, lh) are more and more
(Figs. li, 2) usually arrive within the limits of
able. The seismogramsavailable for the pres- indefinite as the distance in the shadow zone
ent investigationshow that diffractedP waves increases. Beyond an epicentral distance of
having periods of more than about 10 sec about 120ø the existenceof such short-period
IOO
I
I
I
I
I
I
IiO
120
130
140
150
160
+o
enfral d;slance e in degrees
curvefor loga'•' Knopoff
I
T=2 T sec
log a/T observed
•
T_< Isec
+ Isec <T<$sec
P vertical
0 $sec<_T_<9sec P horizontal
1957,April16
•, T>9sec
ha600
1924,,June
km
h=$O km
m=7.2
I00
I
m=7{,
I10 .-•e
I
120
120
1::50
I
140
I
150.-•e
I
160
•
Fig. 3. Logarithmsof ratio of observed
amplitudes
a in micronsof longitudinal
wavesP to periods
T in secondsas function of epicentraldistance• in degreesfor two earthquakes;theoreticalcurves
after $cholte [1956] and after Knopo• and others [1959], and curve from model experiment after
Rykunov[1957].Calculated
curvesarebasedon Gutenberg
andRichter[1956]for therespective
magnitudes and depths of loci.
SHADOW OF THE EARTH'S CORE
diffracted P waves can be establishedbeyond
reasonabledoubt rarely earlier than about 15
sec after the calculated time, and they are
usually observable only in records of large
earthquakes.The delay in arrival may well be
related to the delayedmaximum, predictedby
the theory for a rectangularpulse.
The beginningof diffracted S 'waves is disturbed by the motion preceding it. In the
shocksinvestigatedhere no S waves could be
identified beyond a distance of 125ø; the observed travel
times of the diffracted
TABLE 1. Approximate Ground Amplitudes a
and Periods T for Horizontal ComponentsPH of
P and SH of S for the Earthquake on June 26,
1924, 01 37 UT, Based on Records Reproduced
by Macelwave [1930] or Krumbach [1931]
PH
Station
S waves
(Figures lk and 1/) agreewithin about ñ15
sec with
the values calculated
1017
from the extra-
polated travel-time curve for S.
For the shocksof June 26, 1924, and April
16, 1957, trace amplitudes of diffracted P and
S waveshave beenmeasuredand groundamplitudes have been calculatedas far as possible.
#, deg a, •
Rio de Janeiro
98.7
Hyderabad
98.8 . ?
15q-
SH
T,
T,
see a, •
see
14
60
15
?
25
12
5
10
107
Pasadena
114.8
0.5?
7
Victoria, B.C.
123.4
1.2
10
4?
Sverdlovsk
Ottawa
Uccle
Paris
De Bilt
Oxford
140.6
145.4
163.4
163.6
163.6
167.0
0.5
0.3
0.2
0.4
0.3
0.17
18
107
15
14
15
10
......
......
......
......
......
......
The instrumental constants for 1924 were taken
from Krumbach [1931] for the mechanicalinstrumentsand from McComb and West [1931]
for the Galitzin instruments.For 1957, the constants were taken from data furnishedby the
stations or from station bulletins for 1957. The
resulting ground amplitudesa and periods T
are listed in Tables I and 2; the logarithmsof
a/T for P are plotted in Figure 3 and thosefor
TABLE 2. Approximate
GroundAmplitudes
a andSPeriods
T of theEarthquake
onApril 16, 1957,
04 04 UT, Magnitudem -- 7.2, Basedon OriginalRecords.
PZ*
Station
Kiruna
Uppsala
Honolulu
Stuttgart
Tamanrasset
Alger
Kew
0, deg
93.0
93.8
95.7
98.8
102.8
!04.3
a, •
PZ*
T, sec
T, sec
a, •
16
16
14
10
16
6
4
9
7
6
2.7
3.5
6
4
(•:•
'•'
'•'
'•'
•'•
'•'
0.7
1•
2•
3
3
8
......
104.6......
109.3
112.6 6'b27
J: '•'
118.5 _•0.002 -•
1.6
0.8
............
3
2•
Santa Clara
Saskatoon
124.3
124.8
127.5
(•'(•01 ]
Woody, Cal.
127.8
O.002
Pasadena
a, •
9
10
Toledo
Coimbra
Victoria,
B.C.
Eureka,
Nev.
T, sec
0.7
106.4
107.7
108.1
SH•
1.4
......
Relizane
Rathfarnham
Sitka
a, •
PH•
128.4 _•0.5
7
8
"•« "• •'1'¬127••' •'
..................
............
'•'
'•'
(•'•5
'•'
. ...........
. .....
. .....
..................
•a
4
T, sec
•) •)lq- '•
/)ii3
•'
......
............
'•
'•-
'•-
•'
I q2 q-
6
7
............
. ...........
* PZ is the vertical componentof P.
$ PH and SH are the horizontalcomponents
of P and S.
J:Magnification
of instrumentbasedon ratio of traceamplitudeto calculated
groundamplitudeof PP
for shockat depth of 600 km and m -- 7.2.
1018
B. GUTENBERG
alculated
for
h=600km,
m=7.2
+
+
Calc•lai
SH
'•h
'SH
rn=7•,,,, ,,"
S horizontal
*•/,
S horizontal
+
+
-0.5
+
+
1924, June 26'•
1957, April 16%
I
h=600km+
•
m=7.8,105T•15sec
%+•
I
h=30km
•
m=7.2
om
---• Epicentroldistance,
-
degr.
I10
I00
120
20.•picentral
distance,
I00
I10 degr.•%
120• +
1.0
I
I
Fig. 4. Logarithmsof observedvalues a/T for transversewaves; details as in Figure 3.
SI-I are plotted in Figure 4. It shouldbe noted
that
the
various
authors
of
the
theoretical
curvesreproducedin Figure 3b assumeddifferent distancesfor the beginningof the shadow
zone. On the other hand, Gutenberg [1957,
p. 319] found that in P waves at distancesof
less than 100ø 'the ratio a/T is approximately
the same only for periods T of less than 8ñ
seconds,but it decreasesappreciably if the
periodsincreasebeyond 10ñ seconds;for peTABLE 3. Approximate Values of the Exponential Factor ] (Equation 2) for the Decrease
of Diffracted Longitudinal Waves in the Shadow
Zone
of the
Earth's
Core
T, sec f/megameter
1104-
0.4
704-
0.6
20
0.94-
12
8
6
2•
0.8
0.8
1.2
1.2
_•1
1.9
Source
Rykunov[1957] from model
experiments
Rykunov[1957]from model
experiments
Pasadena,Table 2, discussed
in text
Observed,Figure 3b, Table 1
Theoretical, Scholte[1956]
Observed,Figure 3a, Table 2
Theoretical, Knopoffand
others[1959]
Observed,Figure 3a, Table 2
riodsof 20 seconds
it is only roughly• of the
correspondingvalue for waves with periods of
4 seconds.' This has to be considered in com-
paring the observedvalues of a/T for T •
sec for
P waves with
the theoretical
8
curves.
From such a comparisonit follows that the
observeddecreasein amplitudesof P in the
shadow zone is approximately the same as
the decrease calculated for waves diffracted
at
the core boundary.
For a numerical comparisonwe may assume
that the amplitudesin the shadowzone decrease
from diffraction by a factor e-•" with the distance D in megameters. If 0• and 02 are the
angulardistances
in degreesof two pointsin the
shadowzones,
D -- (•
-- •)/9
megameters
(1)
Considering that the waves arrive on zones
which decreaseunder otherwise equal circumstancesas the anticenter (• -- 180ø) is approached,we assumethat
_f= .2_•
log
a2
sin
O•.permegsmeter
a•
sin
0•
(2)
We find from Figure 3b that f -- 0.8 for longitudinal waves having periodsof about 12 sec,
SHADOW
OF THE
EARTH'S
CORE
1019
andfromFigure
3athaty - 1.9forlongitudi-1959, p.
nal waveshaving periodsof about 1 sec.As has
been mentioned, the value of &T for longitudinal waves having periodsof 20 sec at distances less than 100ø is roughly one-fourth of
thosehaving periodsof lessthan about 12 sec.
Similarly, recordsof PH at Kew give for &T
about 0.2 for T --
7 sec and about 0.06 for
T ---- 12 sec. Consequently,the observationat
Pasadena,• -- 128.4ø (Table 2, Figure li and
3a) would give roughly • -- 0.9 for P waves
having periods of about 20 sec. Effects of the
focal depth of 600 km on amplitudes(and distances) are within the limits of errors involved
in the presentresearch.
A comparisonof the values of • for diffracted longitudinal waves derived from observations with those from theoretical
results in
90] applied Airy's theory of diffraction and found that waveswith periodsof 1
sec shouldbe traceableback only a few degrees,and thosewith periodsof 10 secnot over
about 15ø.
l•EFERENCES
Friedlander, F. G., Diffraction of pulses by a circular cylinder, Communications on Pure and
Appl. Math., 7, 705-732,1954.
Gutenberg,B., Ueber ErdbebenwellenVIIA,Nachr.
Ges. Wiss.G6ttingen,Math.-Phys. KI., 166-218,
1914.
Gutenberg, B., Bearbeitung von Aufzeichnungen
einiger Weltbeben,Abhandl. senckenberg.natur[orsch.Ges.,40, 57-88, 1925a.
Gutenberg, B., Der Au[bau der Erde, Gebr. Borntraeger, 168 pp., 1925b.
Gutenberg, B., Spectrum of P and S in records of
distant earthquakes, Z. Geophys., 23, 316-319,
Table 3 shows good agreement. Within the
1957.
limits of the periodsT used,all data in Table 3 Gutenberg, B., Caustics produced by waves
through the earth's core, Geophys. J., 1, 238can be representedfairly well by • = 1.7 248, 1958.
0.6 log T. This would indicate that the equaGutenberg,B., Physicso• the earth'sinterior, Acation for the decreasein amplitudesof P in
demic Press, New York, 240 pp., 1959a.
the shadowzone containsa factor of roughly Gutenberg, B., Earthquake waves reflected at the
T"/,.
inside of the core boundary, J. Geophys. Re-
search, 64, 1503-1508,1959b.
Summarizing, we may conclude that the
Gutenberg, B., and C. F. Richter, Materials for
wavesobservedin the shadowzoneof P beyond
the study of deep-focusearthquakes,Bull. Seis.
0 ----103ø exhibit the propertiesto be expected
•oc. Am., •6, 341-390, 1936.
theoretically for diffraction at the core bound- Gutenberg, B., and C. F. Richter, On seismic
waves IV, Getlands Beitr. Geophys.,54, 94-136,
ary. There is no indicationthat an assumed
de-
creaseof the velocityin the lowestportionof
the mantle
could lead to the observed 'dif-
racted'wavesat distancesbeyond•'bout 110ø.
On the other hand, available calculationsindicate that P wavesarriving at distancesof 90ø
cannothave gonedeepenoughto grazethe core.
A constantvelocity or one decreasing
by about
0.1 km/secin the portionof the mantle,roughly
100 km thick, just above the core would account for the nearly straight line travel-time
curves and the decreasingamplitudesof P
wavesat epicentrM distancesbetweenabout 90ø
and slightly over 100ø.
1939.
Gutenberg,B., and C. F. Richter, Magnitude and
energy of earthquakes,Ann. geofis. Rome, 9,
1-15, 1956.
Jeffreys,14., The times of P, S and SKS, and the
velocities of P and S, Monthly Notices Roy.
Astron. Soc., Geophys.Suppl., 4, 498-533, 1939a.
Jeffreys,H., The times of the core waves,Monthly
Notices Roy. Astron. Soc., Geophys.Suppl., 4,
548-561, 1939b.
Jeffreys,H., The Earth, 4th ed., Cambridge Univ.
Press,420 pp., 1959.
Knopoff, L., and others, Seismicscatteringproject, Am. Petrol. Inst., Project 59, 4th ann. report, 208 pp., 1959.
Krumbach, G., Die instrumentellen Au]zeichnunThe relativelyslow decreasein amplitudes ge• des Erdbeben vom 26. Juni 1924, VerSff.
f. Erdbebenforsch.,Jena, Heft
of the wavesdiffractedat the core boundary Reichsanstalt
16b, 9 pp., 24 pl., 1931.
and arriving at epicentral distancesbetween Krumbach, G., Die Ausbreitung von Erdbeben•'bout 103ø and the anticenter contrasts with
wellen iN grossen Herdentfernungen bei dem
Siidseebebenvom 26. Juni 1924, VerSff. Reichthe muchmore rapid decrease
of the amplisanstalt f. Erdbebenforsch.,Jena, Heft 16a, 68
tudesnear caustics[Gutenberg,1958]. For this
pp., 1934.
decrease,but not for the decreasein the shadow Lehmann, I., On the shadow zone of the earth's
of the earth'score, Jeffreys[1939b,p. 553;
core,Bull. Seism.Soc.Am., 43, 291-306,1953.
1020
B. GUTENBERG
Lehmann, I., On amplitudes of P near the shadow
zone, Ann. geofis. Rome, 11, 153-156, 1958.
Lehmann, I., On amplitudes of P near the shadow
zone, Chronique de I'UGGI, no. •2, p. 230, Abstract with discussion,1959.
Macelwane, J. B., The South Pacific earthquake
on June 26, 1924, Getlands Beitr. Geophys.,28,
165-227, 1930.
McComb, H. E., and C. I. West, List of seismological stations of the world, Bull. Natl. Research Council U.S., no. 82, 119 pp., 1931.
Rykunov, N., A study of the character of diminution of amplitudes of waves in the shadowzone
of the earth model, Bull. Acad. Sci. USSR.,
Geophys. Set., no. 10, 73-77, 1957; translated
from Izvest. Akad. Nauk SSSR, set. geofiz.,no.
10, 1262-1265, 1957.
Scholte, J. G. J., A mathematical discussionof
seismic waves diffracted by the earth's core,
Assoc. Seism. et Phys. Int. de la Terre, Rome,
1956, Comptes rend., no. 11, 57-58, 1955; Abstract and discussion.
Scholte, J. G. J., On seismicwaves in a spherical
earth, Koninkl. Ned. Meteorol. Inst., no. 102-65,
55 pp., 1956.
(Manuscript received November 19, 1959.)