-
7/28/2019 Parnell 54
1/32
SOXATOTTPING BY PHYSICAL ANTHROPONETRPR. W. ARNELL
T he W n m f o d Hosp i ta l , O x f o r dTWO F I G U R E S
I n 1940 Sheldon, Stevens and Tucker introduced theirmetliod of
somatotyping by inspection of standard photo-graphs. Their work has
provided an immense stimulus toanthropologists. While fully
recognizing the great value ofthis contribution there are certain
practical difficulties whichhave hindered the more general adoption
of their method.Among thcse difficulties are the following:
1. Standards of somatotype dominance are subjectivelydetermined.
This means that where the metric method isused and the labor of
taking 17 photographic measurements,calculating the corresponding
ratio indices and looking upthe most fitting somatotype in tables
has been correctly com-pleted, the result may still be wrong if the
original choice ofdominance was incorrect. It is believed that the
long metricmethod of typing recommended in Varieties of
HumanPhysiqueby Sheldon and his colleagues has largely falleninto
disuse and this is no doubt partly due to the great deal
I a m most grateful to my personal assistants, to Nrs, C. C.
Standley w horneasnrcil iiiost of the photographs aiid carried out
the photometric typing ac-cording to my photoscopic estimates, and
t o Mrs. S. C. McIntosh for statisticalhelp especially in compiling
the deviation tables. Dr. J. M. Tanner introducedme to
soni:itotyping aiid himself measured and photographed many of the
Oxfordundcrgrndu:ites. I h . R. IT. Bolton, Senior University
Medical Officer, kindlygranted me permission to carry out the
anthropometric survey of first year stu-dents : i t Birniing1i:im
in the course of their routine overhaul. The first pa r t ofthe
work was iuiderttiken while I mas Student Health Physician at the
Instituteof Social Medicine, Oxford, the la tter p ar t while in
receipt of a further grantfrom the Snffield Fonndntion to undertake
research in the constitutional aspectsof psyr1ii:itric riwdiriiie
:it t l w Warneford Hospital, Oxford.
20 9
-
7/28/2019 Parnell 54
2/32
210 R. W. PARNELLof labor involved, amounting in my experience
to not lessthan one hours work for each subject. If this is so then
animportant prop of objectivity has been removed and therereimins
insufficient guarantee against shifting standards ofrating. I n due
course a more complete reference book ofsomatotype photographs inay
become available for generaluse. This will help, but agreement as
to a given somatotypewill still depend 011 personal interpretation
of visual impres-sions, not upon measurement. If somatotypers agree
it willmean that they have learnt to sing in harmony, but their
songdoes not thereby become a science, it remains an art.
Sheldon,Hart1 and McDermott (51) emphasize the skill required
inanthroposcopic somatotyping and say omatotyping cannotbegin and
end with milliiiietcrs. It is agreed that the photo-graphic record
provides information in more assimilable formthan any large iiumber
of bare measurements, thereforesomatotyping cannot end with
millimeters, but it inay rca-sonably begin with measurement and it
is hoped to show herethat physical aiitliropoiiietry call provide a
useful degree ofscientific objectivity as a preliminary guide to
somatotypists,though inspection of the photograph may lead to some
slightsubsequent revision of this preliminary estimate.2. The
second difficulty is that objection to being photo-graphed in the
nude niay render the somatotyped sampleunrepresentative of a
population chosen for study. This ap-plies more particularly in
somatotyping women.3. A third difficulty not infrequently
encountered is tofind accommodation for the 10 meter camera-subject
distancerecommended by Tanner and Weiner together with the costof
photographic equipment, development and standard en-largement of
photographs.
The purpose of this paper is to describe a short
physicalaiithropometric method, which can be used during
clinicalinterview for the following purposes :
(a ) To provide objective guidance as to dominance ofsomatotype
in healthy persons.
-
7/28/2019 Parnell 54
3/32
S O M A T O T Y PI X G B Y A K T H H O P O M E T R Y 211(b) To
estimate the Sheldonian somatotype objectively
and as accurately at least as the agreement achieved
betweenexperts at photoscopic somatotyping.( c ) To makc an
estimate of womens somatotype possiblealthough in the absence of a
published reference file of photo-
graphs this estimate cannot yet be checked by
photoscopicstandards.
(d) To reduce the cost, labor, delay and other handicapsinherent
in photometric methods.D E V I A T J O S CH-YRT PROFTLE O F
PHYSIQUE
The method of estiniating dominance depends primarilyon what
will be described as the Standard Deviation Chart.On this chart
(see table 1) standard scales are shown forheight, weight,
height/vweight (hereinafter referred to asthe ponderal index), for
two bone sizes, the bicondylar ineas-urements of humerus and femur,
for two muscle girths, namelythat over the tensed biceps with fully
flexed elbow and thecalf girth standing, and lastly for three
skinfold measure-ments of subcutaneous fa t and the total of these
three f atmeasurements. A discriminant function scale of
androgynyas described by Tanner (51) is also included but this is
ofsecondary importance for estimating somatotype. Subsidiaryscales
are used also for biacromial measurement, bi-iliac,chest width and
chest depth, but these too are of secondaryimportance for the main
purpose, though they serve at timesto provide useful supplementary
evidence.
The scales were each plotted around the mean value, withone
column unit equivalent to one-half standard deviation,giving a
13-point scale over The necessary measurementsoccupy about 5
minutes in the taking and by ringing theappropriate measurement for
a given person on the deviation
These are ex ten sive scales which differ in certain respects
from the equalappearing interval scales used in soinntotyping. See
Varieties of HumanPhysique, p. 115. Note that the scales are
extended in the minus directionbeyond the - 3 l ine; this i s to
cover the range of female values, and thefemale means are shown in
boxes.
-
7/28/2019 Parnell 54
4/32
Nme
-__A__n
Smaoy
1917161317.1-i&O-1mI
1726
12115
1791
--_
Cdhen
--__
HW~R~T
-_
~.
ST
SC
252322202826
141618101214
lLg(n)
Weg1s
HW.Ro
Io:Hmu
(
Fm
Me:Bc
_
(cm~C
--
F:S
su
a
(nmS
suaa
S
o
c
-5
-44
-4
-31
-3
-2i
-2
-18
-1
-&
Y
+3
+I+I&+2
+2&+3
555861646663.8I60163653
7
72i2476797173
7
8
91
1f21
1
11
111
1
2
66
15171013161811141611C12151710131518
505.2
54555759I61163__64-
.810-7173757778
77
7981838587
89I
911
93959719911131
1719
1829
2022224
25126I272930'~-____-
313233343536
2222
222232323333g3T33133334444444
__-
.
.
5060
70T15
15T,Fp52300+
-
3040
50658010101010202030+
3545
5060759015151010202030
-
7/28/2019 Parnell 54
5/32
SOMATOTTPING BY A N T H R O P O M E T R Y 213chart, a profile is
outlined which reveals the main physicalcharacteristics at a
glance. The profile may be read on sightand an opinion formed as to
somatotype dominance. Armedwith this and Sheldon's somatotype table
for each ponderalindex, a fairly accurate estimate of somatotype
may be ob-tained, and there is the distinct advantage that the char
t maybe employed to provide an estimate of somatotype during
theclinical interview.
For subsequent more exact quantification or to guide be-ginners
in interpretation of the profile, a set of tables issupplied in the
appendix which give estimates of the direc-t ion of
endomorphic-mesomorphic dominance correspondingto each bone and
girth measurement, for each height andponderal index. These tables
were derived from the measure-ments of 105 undergraduates at
Oxford, 1948 to 1951, and 508first year students at Birmingham,
1952. When constructingthe tables, age limits were set from 17 to
24 years inclusive.The Oxford series had been somatotyped using
Sheldon'slong photometric method; in addition 283 had also been
in-dependently typed photoscopically (78 of them by Dr. C.
T1'.Dupertuis) as reported elsewhere by Tanner ( ' 5 2 ) .
V A L I D I T Y O F RESULTSDeviatiolz t a b l e t y p i n g c
omp a r e d with t h e
p h o t om e t r i c methodEstimates of somatotype obtained by
physical anthro-
pornetry have been compared with results obtained usingSheldon'8
long photometric method. For this comparisononly men aged 16-20
have been included according to thelimits within which Sheldon's
tables were standardized. Therewere 151 men of that age and table 2
summarizes the differ-ciices. A plus sign implies that deviation
table ratings werehigher.It will be seen that in 90.0% of cases
ratings mere correctto half a unit. This compares well with the
agreement be-tween experts using the photoscopic method (Tanner,
'52).
-
7/28/2019 Parnell 54
6/32
214 n. W. PARNELLThat the agreement is not actually closer could
be due inpart to incorrect photometric typing, since the essential
pre-liminary photoscopic estimates of dominance might havebeen
wrong.
TABLE 2Am o u n t s by whicli deviat ion table est imates of 154
Oxford m e n d i f f e r e d f r o m
ra t ings based o n Sheldon's photoniet i ic tablesMEAX+' +'
IFFERENCEIFFEBENCES -13 - 1 - 4
Eiidomorphy 1 11 58 5 7 23 4 - .169Mesomorphy 1 17 40 59 32 5 -
.073Ec tomorphy 0 2 27 81 39 5 +0.058T o t a l 110. r a t in gs 2
30 125 197 94 14 462T o ta l pe r ce nt 0.4 6.5 27.1 42.6 20.3 3.0
99.9
TABLE 3Differences between deviat ion table est imates and
photoscopic rat ings
in $82 Oxford m enY E A NDIBPERmNCES - 1 b - 1 - 1 0 + f +' I h
DIFFERENCE
Endomorphy 1 25 82 97 64 12 1 - .08Mesomorphy 0 7 44 95 98 32 6
+0.22Ec tomorphy 0 1 0 55 146 57 13 1 +0.02To t a l no . r a t i n
g s 1 42 181 338 219 57 8 846T o ta l pe r ce nt 0.1 4.9 21.4 40.0
25.9 6.7 1.0 100.0
I)e,iTiation table t y p i n g co m p a r ed with
anthroposcopyDeviation table estimates have also been compared
withphotoscopic ratings (see table 3 ) . The subjects for this
com-parison numbered 282 from the series of Oxford undergradu-ates
reported by Tanner ( ' 52 ) .I t will be seen that in 87.3% of
cases the ratings werecorrect to half a unit. There was a tendency,
however, inthe photoscopic method to rate mesomorphy lower hy
roughlyone-fifth of a unit.
-
7/28/2019 Parnell 54
7/32
SOMATOTYPING BY ARTHROPONIETRY 215DETAILS O F T H E METHOD
Construction of the standard scales on the deviation chartwas
straightforward where the frequency distribution wassufficiently
natural or Gaussian in shape. This was generallyso, but there were
two exceptions. The ponderal index dis-tribution showed some
skewness. Though this ratio correlatesvery closely with ectomorphy
the standard scale does notcorrespond precisely with Sheldons scale
of ectomorphy andin order to make a provisional estimate of
ectomorphy fromthe ponderal index corresponding values are included
in thelowest section of the deviation chart. The values are
derivedfroni table 23 in Varieties of Human Physique and
aprovisional estimate may be expected in young men aged16 to 20 to
be within plus or minus half a point of the truevalue, with only
rare exceptions where the provisional * esti-mate may be as much as
one unit out. The skinfold measure-inents of subcutaneous fat also
showed a skewed distribution.To relieve the skewness fat
measurements were converted tologarithmic scales and the same thing
was done with the totalof the three fat measurements. Once the
scale had been con-structed, however, the appropriate number could
be ringedas for the other individual measurements and no
referenceto logarithmic tables is necessary when using the
deviationchart.
Method of taki i tg the measurewze&H e i g h t was recorded
in inches. The subject stands backto a wall scale, takes a deep
breath and stretches up to maxi-
mal height, his heels remaining in contact with the ground.W e i
g h t is recorded to the nearest pound, for the Oxford
series this was without clothes, for the Birmingham seriesa pair
of pants and socks were permitted, weight approxi-mately 8 0 2
.Bone nzensuremeMts (in centimeters). The distance betweenmedian
and lateral epicondyles of the humerus was takenand secondly the
distance between median and lateral epi-condyles of the femur.
Engineers steel calipers fitted with
-
7/28/2019 Parnell 54
8/32
216 R. W. PARNELLVernier scale were used for the femur but
ordinary steeloutside curved calipers fitted with screw adjustment
do equallywell and a re preferable f or the elbow. The points of
thecalipers with the measurers index finger alongside a re
placedfirmly against the tips of each epicondyle and the
subjecthimself tightens the screw. A steel centimeter rule
allowsthe distance between caliper points to be measured to 0.5
mni.The caliper points are slightly blunted with a file and theskin
is eased away with the forefinger to prevent scratchingwhen the
calipers ar e removed. Alternatively the spl it screwmay be
released half a tu rn while the calipers a re withdrawnand
subsequently tightened by the same amount.
M u s c l e g i r t h mrnsziremeizts (i n centimeters). Biceps
gir thwas taken with a highly flexible steel tape i n light contact
withthe skin over a tensely contracted biceps with the elbow
fullyflexed. Calf gi rth taken with the subject standing erect,
thelegs almost touching and the tape in light contact with theskin.
The maximal gi rth was recorded.
S k in fo ld mcasurenie iz t s of s u b c u t a n e o u s f a t
(in millimeters)mere recorded with modified Franzen subcutaneous
tissuecalipers at three sites:
1 . S u b s c a p u l a r . The skinfold was raised with the
thumband forefinger of the left hand over the angle of the
scapula,the skinfold running downwards in the direction of the
ribs.The subjects a rm hangs by his side. The skinfold shouldnot be
held too tightly because it is tender when pinched.
2. S u p r a i l i a c . The skinfold is raised as before with
theleft hand in a positioii one to two inches above the
anteriorsuperior iliac spine, and the fold is raised so that it
runs inthe direction of the intercostal nerves.
3. Over t r i ceps musc l e . Halfway between the acromionand
the olecranon on the posterior aspect of the arm. Carewas taken to
make sure no muscle fibers were included; incase of doubt if the
subject locks his elbow momentarily themuscle fibers will withdraw
from the fold. Th e elbow shouldnot be held locked for this par tly
te thers the skin.
-
7/28/2019 Parnell 54
9/32
SOMATOTYPING BY ANTHROPOMETRY 217Biacroiiaiul width
(centimeters). For this it is important
to make certain that the shoulder muscles are relaxed andthat
the shoulder girdle is not braced back or upwards, neithershould
the shoulders be rounded too far forward. Comparisonof the results
obtained using a pelvinieter and an anthro-pomenter showed that the
pelvimeter with its diminished scalewas not a fully satisfactory
instrument. A suitable instru-ment has been made at very moderate
cost by fitting arms toa standard 50cm steel rule. The arms are
pressed firmlyagainst the outer aspect of each acroniial
process.Bi-ilim w i d t h (centimeters) is taken between the
outeraspects of the iliac crest using firm pressure against the
bone.
TABLE 4Correlation coefic iests betw een individual s t inf old
measurements
and the mirn of all three410 164O X F O R D MEN O X F O R D
WOMENA G E D 17-24 A G E D 17-23
Total fat and subscapular fatTotal fat and suprailiac fatTotal
fat a d riceps region fat
r =0.97r =0.93r =0.84
r =0.82r =0.81r =0.82
(:hest width a n d chest d e p t h (centimeters) were taken
withthe thorax midway between full inspiration and expiration.The
arms of the pelvimeter were held horizontally level withthe
greatest width or depth. This level varied quite widely.Aiiatomical
precision was foregone with the object of indi-cating the
relationship between maximal antero-posterior andmaxinial lateral
dcvelopmeiit of the thorax.
The t h r e e f a t m e a s u r e m e n t s a s a n i n d i c a
ti o nof t h e t o ta l s u b c u t a n eo u s f a t
It is necessary to consider how far these three
skinfoldmeasurements may be taken to indicate the total amount
ofsubcutaneous fat in the body. The first step taken was
tocorrelate each of the three subcutaneous measurements withthe sum
of all three. The results are given in table 4. These
-
7/28/2019 Parnell 54
10/32
218 R. W. PARTSELLresults are encouraging and perhaps a little
surprising, forit seemed clear after only short clinical experience
that sitesof adiposity vary to some extent from one individual
toanother. In order to evaluate the sum of the three fat
meas-urements used in the survey as an indication of general
sub-cutaneous fat in the body it was desirable to correlate
thistotal with totals obtained using a much larger number
ofmeasurements.
Edwards ( '50) published interesting observations on
thedistribution of subcutaneous fat. He had measured 53
sitesincluding the three used in this survey and he has been
kindenough to allow me to use his measurements. Edwards' sub-jects
were 24 obese women but 48 sets of measurements wereused since each
person was measured before and after 28 lbs.or more reduction in
weight. The coefficient of correlationbetween the sum of
measurements at 53 sites and the sum atthe three sites used in this
survey showed very close agree-ment, 0.99. This high degree of
correlation in women, amongwhom variation of individual
measurements is greater thanin men, made it possible to proceed
with a fair degree ofconfidence that the sum of the particular
three measurementschosen was a fairly good indication of the total
subcutaneousfat in the body.
These remarks do not imply that individual variation doesnot
occur in the pattern of fat storage. It was clear in table4 hat in
men fat in the region of triceps departed more oftenfrom the
average pattern. A large amount of fat over thetriceps muscle is in
fact a feminine feature and in men highfat measurement in this
region relative to others suggestsgyna~idromorphy.~n mesomorphs the
fat measurement withthe highest standard score is commonly the
subscapular one;in endomorphs the suprailiac.
The total of the three fat measurements will be used inthe
deviation chart profile to obtain a provisional estimate
* In the Oxford series, excluding central somatotypes, 61% of
endomorphic-ectomorphs had their highest fat standard score in the
triceps region, but amongmesomorphs this feature was only present
in 22%.
-
7/28/2019 Parnell 54
11/32
S O M A T O T Y P I N G BY A N T H R O P O M E T R Y 219of
endomorphy. It is recognized that this ignores other im-portant
anatomical characteristics of endomorphs, for ex-ample their
relatively small bone structure, but this featureis directly
visible on the deviation chart.O n the last line of the deviation
chart there will be founda provisional estimate of endomorphy
corresponding to thetotal fat score in the same column above it.
These provisionalestimates represent average endomorphy figures in
the Ox-ford series (photometric rat ings) corresponding to the
totalfat score. The ultimate justification for using total fat
meas-urements a s an indication of endomorphy may be judgedfrom the
closeness of the final estimates obtained by thismethod and by
expert anthroposcopic typing.
The profi lesThe varieties of somatotype dominance are found on
Shel-
dons chart (fig. l)on which incidentally the distribution of405
Oxford men has been plotted, as derived by deviationtable
typing.The next step is to describe the main characteristics
ofprofiles corresponding to each somatotype dominance. Fora person
of about average height, that is for someone whoseheight lies
within approximately three inches of the mean,figure 2 illustrates
the relationship between height, bone,muscle girth and subcutaneous
fat standard scores.
In the diagram, B is taken as the average of the twobone
standard scores, M is the average of the two musclegirth scores and
a cross is used to indicate an average valuefor bone and muscle
development. A straight line has beendrawn connecting the height
and total fat standard scores.
In endomorphs and endomorphic-mesomorphs this lineHF lies in the
direction of the French accent grave(top left to bottom right).
This was found to be true of eventhe tallest endomorphs in the
series.In ectomorphs and ectomorphic-mesomorphs the line HF
runs in the direction of the French accent aigu (top right
-
7/28/2019 Parnell 54
12/32
220 R. W. PARNELLto bottom left). The only exception t o this is
in short ecto-morphs but here the ponderal index will save
confusion sinceat this age all ectomorphs, including those with
shared pri-mary dominance, have a ratio exceeding 13.25.
A SCHEMAIIC TW O DIMfNSIONAL PROltCTION OF THE T H t O R t l l C
I L SP4TIAL RtLITIONSHIPS 4MONG TH t K N O W N SOMATOTIPIS
Fig. 1 The soiliatotype distribution of 405 Oxford men
undergraduates.The line HE is vertical in the soniatotypes 444 of
aver-
age height. It is vertical also in endomorph-ectomorphs
(434,424) and primary mesomorphs (343,353,262) when
secondarydominance is shared.
Primary mesomorphic dominance is present where the B M average
point marked with a cross lies to the rightof both the height
standard score H and the HF line.
-
7/28/2019 Parnell 54
13/32
SOMATOTYPING BY A N T H R O P O M E T R Y 221
ENDOMORPH EnomorphieBX
ENDOMORPH/\ENDOMORPH
Mesamomhie"aEndomorph = Mesomorph
H MESOMORPH Endomorphic9
/ x "ESOMORPH Enomorphicd M
d ECTOMORPH Meso-tphieM
ECIDMORPH/ =
MECTOMORPH Eodomorpliic
F i g . 2 Diagram illustrating the relationship of height (H),
bone (B), niusclegirth ( M ) and subcutaneous fa t (F) standard
scores 011 the deviation chart accord-in g to dominance of
somatotype in a person of about average height.
-
7/28/2019 Parnell 54
14/32
222 R . W. PARNELLhlesopenia (that is mesomorphy 3 or less) is
present where
the BhI average point lies to the left of the HF line.The layout
of the standard deviation chart is such that thehorizontal line on
which the BM average is marked lieshalfway between the height and
total fa t lines. When theBX cross falls on the HF line a rating of
mesomorphy4 is found. As a further rough guide, if the Bhi cross
isdiscovered to the right of the HF line half a point inmesomorphy
is gained for each column passed. Conversely,mesopenia may be
roughly estimated by half point reductionsof mesomorphy for each
column lying between the midpointof the HF line and the BM average
point to its left.There is an exception. I n endomorph-mesomorphs
the pri-mary dominance is shared but the girth measurements
whichhere contain a fa ir proportion of fa t give a slightly
exagger-ated impression of mesomorphy. Allowance has to be made
forthis, the actual amount depending on the total fat estimate.A
reduction in mesomorphy by half a point is usual butreductions of
one unit may be necessary in the presence ofobesity.
One further explanatory note to the diagram
concernsmesomorphic-ectomorphs. BPII*profiles relate to the
meso-morphic-ectomorphs with low mesomorphy, somatotypes 235,236,
136, 126 and 127. BW2 points lying to the right ofthe H F line
relate to somatotypes 145, 245 and 345 wheremesomorphy is rated
higher.S o far these empirical instructions appear perhaps a
littlecomplicated. A momentary pause in which to consider
theunderlying reasons for this procedure may therefore be
anadvantage. The first point is that, in the process of
Sheldoniansomatotyping, size as expressed in terms of height is
excluded ;the main concentration of interest is on body shape,
whichis expressed in body proportions. Height is therefore takenas
the starting point for the guide line HF. It must fur-ther be
remembered that endomorphy and mesomorphy aredefined as mutually
exclusive components. If F is ac-cepted as the other reference
poiiit for the guide line, the
-
7/28/2019 Parnell 54
15/32
S O M A T O T Y P I N G B Y ANTHROPOMETRY 223mesornorphy
estimate based on the relationship of bone andmuscle measurements
to this guide line will be relativelygreat where F has a low value;
conversely it will be rela-tively low where there is a large amount
of fat and the guideline HF thereby moves to the right.
The last stage in estimating somatotype is to select themost
suitable type with appropriate ponderal index andcomponent
dominance from Sheldons set of tables. This stagemay or may not be
completed during clinical interview ac-cording to choice.
Fu r t h er notes on profile interpretat ionWhile it is
understood that suitable ponderal index-somato-
type tables for each 5-year age group from 18 to 63 willshortly
be published, it must be remembered that the s tandarddeviation
chart is standardized for 18- to 24-year-old men.Although the bone
measurements do not change, the devia-tion chart remains unsuitable
for older persons unless appro-priate allowance for age changes in
muscle girth and subcu-taneous fat can be made. Further reference
is made to thisin a later section.
Comparatively low bone standard scores have been noticedin
endomorphs, but skinfold measurements have been usedas the main
guide to endoniorphy. Fat measurements mayclearly be affected to a
considerable degree by environmentalinfluence, and the same is
true, though to a smaller extent,of the girth measurements used in
estimating mesomorphy.It is likely therefore that the results only
approximate thesomatotype, where it is known that the subject is
neithergreatly overweight nor wasted. Endomorphy and meso-morphy
have not themselves been measured, if indeed theyever can be
precisely in terms of their original definition.
as relative predominance ofmuscle, bone and connective tissue,
that is of tissues derivedfrom the mesoderm or embryonic layer.
Evidence from the
hfesomorphy was defined
Varieties of Human Physique, p. 5.
-
7/28/2019 Parnell 54
16/32
224 R. W. PARNELLdeviation chart profiles shows that although
bone and inusclefollow fair ly consistent patterns by dominance of
somatotype,they nevertheless vary somewhat independently of one
an-other.It has already been mentioned that the last stage of
theprocedure just described for estimating somatotype involvesa
change from body proportions based on the extensivescales of the
deviation chart to the 7-point equal appearinginterval scales of
somatotype components. Practically, theadjustment required in the
middle part of the scales is small,seldom if ever exceeding a half
point, and this is fortunatebecause central somatotypes often
present difficulty photo-scopically. Towards the ends of the scales
it is sometimesless easy to interpret a profile correctly, and
experience isnecessary t o read the right answer at sight, but
extremephysiques are less difficult to recognize photoscopically.A
further difficulty arises from the point already men-tioned that
fat contributes to gir th measurements. In ecto-morphs there is
less than average fa t to contribute to musclegirth and
consequently the smaller girth measurements maylead to
underestimates of mesomorphy. But among endo-morphs and many
endomorphic-mesomorphs there is morethan average fat, and
mesomorphy is apt to be overestimatedin consequence, a 443 for
example being incorrectly inter-preted as a 4433. The effect of
this was not fully appre-ciated when first attempts were made to
sight-read the profilesand the result was to rotate the whole
distributioii slightlyclockwise.
Objec t ive es t im a te of dominance by tab lesPartly to
correct this rotation and further to make due
allowance for the effect of height on other measurements,tables
(see appendix) were prepared to indicate the balanceof
endomorphic-mesomorphic dominance for each height andpoiideral
index group separately. These tables were con-structed in the
following manner: first the average bone,girth and total fat
measurements were calculated for each
-
7/28/2019 Parnell 54
17/32
SOMATOTYPIXG BY A X T H R O P O M E T R Y 225height and ponderal
index separately ; scales were then builtaround the average
measurements using the same half staiid-ard deviation units as on
the deviation chart. Next the aver-age eiidomorphy and mesomorphy
ratings were calculatedfor each ponderal index group among Oxford
men (see table5). Finally the scales were placed against the
appropriatetop line estimates of endomorphy and mesomorphy, so
thatthe inean measurements and the mean somatotype ratingscame into
the same column. Thus a guide to the balance
ofeiidomorphic-mesomorphic dominance was obtained by com-paring
total fat equivalent in endomorphy with the average
TABLE 5X e a n ra t ings in endon iorphy and mesomorphy f or
each pondera l in de z in
Oxford undergradua tes aged 17-84AVREAGE R A T I N G
PONDERA L 1NDF.X Endomorphy NesomorphyLess t h a n 12.45 4.9
4.0
12.50-12.95 4.1 3.913.00-13.45 3.5 3.813.50-1 3.65 3.0
3.313.70-13.95 2.7 2.714.00+ 1.9 2.0 - -~
estimate of mesomorphy obtained from the 4 measurements,two of
bone and two of muscle girth, which mainly contributeto this
component.Six examples are given below, one from each
ponderaliiides group, to illustrate the use of the tables and a
varietyof somatotype dominance.1 . Height 73.6 ins. Weight 134 lbs.
Ht./V-Wi. 14.4. Age 21
En d o - r t i r s n est imate1.75 Most f i t t ing somatotype f
ro in2.50 suggested by 2.4/2.0 is:2.90
4/ 8.1.5
2.0 Sheldon 8 tables with dominance
2.4 21$7l . O O 1R 7 Huinerus 6.7 e mFemur 9.G ri ni c M r
Eiceps 27.7cmCalf 35.0 eiii
Total fat 32.0 in m
-
7/28/2019 Parnell 54
18/32
226 R . W. PARNELL2.
3.
4.
5.
6.
Height 74.25 ins. Wei ght 158 lbs. Ht . /V Wt . 13.55. Age
18.
3.00 2.7 Sheldons tab les with dominanceB Humerus 6.8cm 2.25
Femur 9.8cm 2.75X Biceps 29.2 cni2.75alf 34.8 ern
4/10.75
E n d o - ~ i m o s t im a teMost fitting soniatotype
fromsuggested by 4.0/2.7 is:
4.0 4215I
4.25 I14*50 1E 1
-_Total f a t 44.0 mmHeight 69.6ins. Weight 1461bs. Ht./ VWt .
13.2. Ag e 24.B Humerus 6.5 em 2.50Femur 9.8cm M Biceps 30.5cmCalf
34.8 emTotal f a t 20.0 mmHeight 71.4ins. Weight 164Ibs. Ht . / V m
. 13.1. Age 20.B Humerus 7.1cm
Femur 9.5emBiceps 34.3 emM Calf 39.4 em
Total f a t 42.0 mm,Vote: On account of the fat in the girth
measurements, a reduction of th emesomorphic estimate is required
if a close fit is to be found in Sheldonstables.Height 72.9 in.
Weight 184 Ibs. Ht./VWt. 12.8. Age 23.B Humerus 7.2 em 4.25Femur
10.5cm M Biceps 35.6cmCalf 39.4 em
Total fa t 26.0 mmHeight 65.25 ins. Weight 147 lbs. Ht./V-Wt.
12.35. Age 18. Humerus 6.6 emFemur 9.1 emM Biceps 323 emCalf 35.1
em
Bndo-meso es t inmtaMost fitting somatotype from3.75 Sheldon s
tables with dominance4,503.75 suggested by 2.1/3.75 is :
4/15.00--2.1 2414
Endo-meso es t imateMost fitting somatotype from
5.50 4.5 Sheldons tab les with dominance5.50 suggested by
4.0/4.5 is:
4.0 4434/18.00
Endo-naeso estimateMost fitting somatotype from
5.10 4.5 Sheldon s tables with dominance4.00 suggested by
2.6/4.5 is :~
2.6 21524/17.85
Endo-nieso est imateMost fitting somatotype from
5.35 4.1 Sheldon s tables with dornina~~ce3.60 suggested by
5.4/4.1 is :
5.4 54313-4116.45Total f a t 60.0 mm
N o t e : A reduction of one unit in the estimate of mesomorphy
was requiredin order to find a close fit in Sheldons tables, where
the total fa t (60mm)was much above average.
-
7/28/2019 Parnell 54
19/32
SOMATOTYPIX-G B Y A N T H R O P O M E T R Y 227C'ertaim
practical h a t s
The tables give an estimate of the direction of
endoniorphic-mesomorphic dominance, but in adjusting this estimate
tosomatotype ratings experience shows that :
(a) There is a tendency to underestimate mesomorphyamong
mesomorphs with low endomorphy (see examples 3and 5 ) because there
is less than average fat in the girthmeasurements, and furthermore
where eiidomorphy exceedsectomorphy in secondary dominance,
endomorphic influencetends to lower bone standard scores also.(b)
Conversely in mesopenes, especially central or ecto-morphic
mesopenes, bone size standard scores are commonlylarger than girth
scores although the girth scores are tosome degree swollen by fat
content. I n short there is a tend-ency to overestimate mesomorphy
and to underestimate meso-penia. Additional evidence in mesopenes
that the table esti-mate of mesomorphy needs lowering will be found
in femininefeatures such as maximal fat appearing over the triceps,
ora low androgyny score, or again if chest width and depthare both
small in relation to height.It is correct t o stick closely to
Sheldon's ponderal indextables in the final estimate, but it
increases confidence toknow in what direction the slight
adjustments of provisionalendomorphy-mesomorphy estimates are
likely to be neces-sary according to areas on the somatotype chart.
Lastly itis well to remember that in persons more than two
inchesshorter or taller than the inner limit of the end categories
inthe tables provided still further allowance may be necessaryfor
the effects of height.
Dysplasia . With the above points in mind it should bepossible
to estimate the somatotype of a healthy young manwho is neither
over n o r under average weight for his build.The most commoii
difficulty encountered is that connectedwith the presence of
dysplasia, whether this is between theprimary components or in the
form of gynandromorphy.
-
7/28/2019 Parnell 54
20/32
228 11. W. PARNELLHints of both may be present on the deviation
chart, forexample :1. Humerus and biceps scores may be small and
femurand calf scores high or vice versa indicating uneven
develop-ment of arms and legs.
2. X low androgyny score and/or high skinfold score overtriceps
compared with elsewhere are common feminoid fca-tures in men.
In other forms of dysplasia between the limbs and trunkthere may
be no indication on the deviation chart and ithappens every now and
again ( 2 or 3% in the Oxford series)that the dominance of the head
and neck, thorax and abdomenis the reverse of that in the limbs. It
is in such cases thatthe photographic record is particularly
valuable in preventingiiicorrect estimates, quite apart from the
other informationit maF provide. I f therefore, say on grounds of
economy,no photograph is obtained, a special point should be madeof
lookiiig for reversal of dominance during clinical inspec-tion.
Xuch valuable information will certainly be lost ifuse of the
deviation chart and tables is allowed t o replaceclinical o r
photographic inspection. The question is in factoften asked as to
whether somatotyping may be carried outusing the measurements and
tables alone, without even thedeviation chart. Certainly an answer
may be obtained inthis wa7 but the risk of er ror is greater; for
the procedurealthough objective is blind, and if suitable allowance
is tobe made f o r dysplastic anomalies it is better to keep
theprofile in view just as it is better not to relinquish the
photo-graphic record.
Possible sources of criticism. The standard scales derivedfrom
measurement of university students are representativeneither of the
general population nor of other special popu-lations, say in the
armed forces or in industry, whom it maybe desirable to somatotype.
The peculiar advantage of thishighly selected university group is
that on the average the
-
7/28/2019 Parnell 54
21/32
SOMATOTTPISG BS ASTHROPOiMETRY 229Osforcl students are taller
and heavier than any other sectionof the commuiiity and this may be
taken as the physicalexpression of a high tide mark in total
endowment, bothbodily and mental. According to Sheldons scheme the
totalrating of his three components should add up to no morethan 12
in the most highly endowed somatotypes; thus itis reasonable for
somatotype 444 of average hcight to havein addition the average
measurements of the group in whomphysical expression of total
endowment reaches its highestlerel. F o r somatotyping it is not
supremely important thatthe deviation chart scales should represent
any particularpopulation in an absolute sense. Such scales would in
anycase b e liable to revision to allow for secular trends ingrowth
and nutrition. The main concern in somatotyping iswith body shape,
tha t is the sum of body proportions of whichthe profile is the
outline. What scales are used is of lessaccount, the more important
point being that all investigatorsshould use the same
standards.
Apprchension is sometimes felt that mingling two methodsso
essciitially different in their approach as photographicinspection
and physical measurement may lead to confusion,but the deviation
chart should be regarded as an aid in ob-jectification, as a means
of grading small differences the sumof which in widely contrasting
physiques is obvious enoughto the veriest beginner examining
photographs.
Tlic purist may claim that the deviation chart procedure isnot
somatotyping on the grounds for example that endo-nioi*pli:- cannot
be defined solely in terms of subcutaneousfat. Rut skinfold
measurements are a more accurate guideto snlmtaneous fat than
photographic estimates of this fea-tn1.c. Thus it may come ahout in
practice that gains in esact-ness of measurement at any given time
may outweigh lossof a partly speculative hope that the
morphogenotppe canlw assessed more precisely by photographic
inspection. Butn-hi1tcver view is ta1rc.n it w c m s that the same
or at least ac l o w l y allied ohjcctiw is lwing pursued by both
methods,
-
7/28/2019 Parnell 54
22/32
230 R. W. PARSELLotherwise it is hard to understand ~ h yuch
close agreemelitshould obtain between the resu1ts.jl17hat has been
said relates to healthy persons. TI-lieiisoinatotyping ill people
it should help to inquire into thesubjects weight history, weight
at 18 years and subsequentlyif obtainable, the highest known
\\.eight and the subjectsbelief as to his optimum weight. The
standard photograph isof special value to a skillful iiiterpreter
but looseness of theskin, from wasting or dehydration, js unlikely
to be mishedby the physician who carries and uses subcutaneous
tissuecalipers.U s e of i l ze deviat ion chav t f o r es t imat
ing t h e sonuitotylpeof o2der m e n . lT7ithout testing the method
against a s u f i -ciently large saniple of older subjects who have
been somato-typed when young n o absolutely reliable conclusion can
bereached, but there are several points about the deviationchart
which suggest that it s use might be developed for olderage groups.
Fi rs t, height and bone width measurements werem o r e or less
constant from 2 1 to 29 years of age and it isprobable that they
change little up to the age of 50 and pos-sibly afterwards.
Analysis of calf gir th measurements showedan insignificant
increase with age from 17 to 30 years; inthe same period there was
an increase in biceps girth, sig-nificant in the statistical sense,
but amounting to little morethan half a standard deviation, that is
one column on thedeviation chart. In this increase fa t is probably
the chieffactor and the chief age variables to be considered on
thedeviation chart are therefore the weight and skinfold
meas-urements. Their variation may be gauged to some extent,
Hunt ( 52) reriews a rather different system of classification
bring developvdl ~ p rofessor E. A. Hootoii and his associates. In
this the first compoiieiit issimply ealled at, a n d t l ic second
LL~~~u~c i i lar i ty .he third component beingderived from th e
ponder:tl indeu 01 iiidex of n t teuna t ion as he calls it .
Themethod is easier to operate than Sl~eldons terhniqur, but
ratings o h ouslpchange to a great csteiit 11-itli variations in
weight e:iiised by age, diet. exerciseor illness, whereas Slielclou
rontciids. lim\ r ig ht l r is iinccrtain but perhaps morerightlp t
l im in:11iy of! his opponents :idm it, t11:lt his s o m : i t o t
p e ratiagr; are n e a r l yindrpendent of environmental
iiifluence.
-
7/28/2019 Parnell 54
23/32
S O M A T O T Y P I S G BY A N T H R O P O l l E T R Y
231especially if an older subject happens to know what his
weightwas at the age of 18. If not, the average weight
variationwith age is indicated by standard age, height and
weighttables, though these make no allowance either for
individualsoinatotype or for secular trends. If in such tables in
placeof weight for each height and age the appropriate
ponderaliiides is given instead, it appears that whatever the
indi-viduals height a reduction of approximately 0.7 occurs inthe
ponderal indes between the ages of 18 and 50; of thisabout 0.2
occurs in the first 4 years, 0.3 by 27, 0.4 by 32,0.3 by 37 and the
remainder in the last 15 years. Themajority of persons are likely
to require midrange or nearaverage adjustments but it must be
stressed that no exactallo\\rance can be made in this way for
weight changescharacteristic of individual somatotype; for these
Sheldongircs nuiiiei*ous descriptive notes in Varieties of
HumanPhysique.
I IIC. increasing trend in mean values of subcutaneous
fat~~i~asurerneii tsssociated with age in the present series
wassignificant, but under the age of 30 amounted to less thanhalf a
unit on the standard scale for the total of three
fatnicasureinents. The inean value of course conceals whatrcillly
matters here, iianiely what happens t o individuals,but the
standard deviation was also found to increase. Sucha widening of
the scatter might occur if say endomorph-riiesoniorphs mature early
011 one side of the mean whilepersons with low ratings in
cndomorphy reach their matureweight later, some after 30 years of
age, others never gaining\wig11 at all.
Tse of t12e deuiatioii charts for cstinzaliizg t h e s o i m t o
t j j p eof woiizen. No nietric standards have been published for~
~ 0 1 1 1 ~ 1 1o fa r, no irides file or reference atlas of 1mon.n
soinato-types. There being no source of reference it is impossiblet
o tell how far a provisional estimate of somatotype reached1q-
applying certain guiding rules of interpretation t o theclcriation
chart mould in fact approximate to the Sheldoiiiansomatotype.
Doctor Sheldon has, however, been kind enough
r .
-
7/28/2019 Parnell 54
24/32
I:. W. PARSELL
to show me a disti.ibution chart of 1,000 college women inthe
United States. The greatest concentration on this chartis
ininiedintely below somatotypc 343 in the triangle 543, 533,433. S
o w if tho tiwi*age inea~urcmcntsof Oxford womenstudents are
plotted 011 the male deviation chart and theguiding rules are used
to arrivc at a provisional estimateof the average woniaiis
somatotype, it will be found to workout as 43343. This is virtually
identical with the centralconcentration of American women students.
Furthermore,Sheldon uses the same ponderal index scales for women
ashe does for men. I t therefore seems very likely tha t
thepi-iniaq- omponent scales of cndoniorphp, mesomorphy
andc~ctoniorphy approsinia te more closely to male
standards,n-hicli would not be surprising sinco thc earliest work
wasbased o n the study of 4,000 college men.
Stcrwclcrrd sccrlcs f o r w o t i t ~ t ~ .h e question
therefore arisestis to how appropriate such scales arc f o r women
and anyonewho takes the t roub le to plot womens measurements on
amans deviation chart will soon find a number of women
withnieasureiiients bc~~-oiidhe standard scalcs. It is for
thisreason that in the deviation chart illustrated the scaleshare
hccn estcnded clownward to cowr the range of Oxfordwomens
measui*eiiimts, their mean vnlucs being indicatedby hoses.O n t h o
other hand investigators may feel that a deviationchart based on
womeiis measurcnicnts is more appropriatef o r women; they may
prefer dimensions whose nature is moreexactly understood than that
of the so-called primary com-ponents. If these components arc
primary in the biologicalsense that they originate in a common germ
cell for men andwonicn, it might be more logical to choose
measurements half-way hctweeii thc average man and the average
woman asthe clefinitioii of the central somatotype 444. But it is
notso easy to change ships in mid-ocean. The ship of somato-typing
w t i s launched over 13 years ago. A great deal ofwork has alrcndy
hcen done on the correlation of somato-type with tisnits of
temperament, psychosis and psychoneuro-
-
7/28/2019 Parnell 54
25/32
sis. I11 addition if a change is to be recoiiiiiieiided to
womensstaiidards the oiily ones at present arailablc arc based
011the ~iieti~ureiiicntsf college women. Although this
highlyselected group has the one distinct advantage already
men-tioned, it must not be forgotten that it differs physicallyfrom
the general population. #or esaiiiple the average pon-dcral iiidex
at Osforc l was 12.86 and although this is veryclosc to th c figure
for Ainei.ican college women, a poiideraliiidox derived fro111 the
average Iiciglit and average weightof 18-year-old iiisured woiiicii
would bc tihut 13.1.
lroni the quandary thus rctiched it is iicwcssary t o returnt o
tlic lwiniary questioii an d purpose^of this discussion- ana
cieviatiou chart lw U S C ~ o estimate tlic She ldo i i i an
somato-type of woincn ? The oiily ~*easoiial)le I ~ S W C ~ t
present isthat by using a I ~ ~ R I L Seviation chart it may be
possible butit has not pet heen sho\vii how closely this may lie
done es-c ~ > p tii the case of thc average Osfortl If
aninvestigator, 011 the other hand, prefers to typc women
against\\-oiii(biisstand ard measurements, he may use their
deviationc1itii.t and follow rules of iiiterpretatioii of his own.
In thisCNSC it would havc to lie clearly understood that if 110
refcr-c~ncc s made to Sheltlo~is ables thc results obtained wouldlw
qui c distinct fi*oni mid p-obahly not c wn a close
approsi-iiiatioii to Sheldoiiian soinatotype. The final clioicc of
stand-;ii*cls is likely to he settled lip coniparing thc usefulness
ofc;wh iiiethod for. the pal-ticiilar purpose in ricw.
STTJfJI.\RT -\XI) O S C I , I S I O S 81. A niethocl is
c1ewril)ed of cMiniating Rheldonian Fomnto-type in young 1iie11,
aged 17 to 24, by ph;r-sical anthropometry.Siiicc the measui*ements
occupy onlp 5 miiiutes in the takingthe method is suitahlc for
general clinical use, the result beinga\-ailal)lc i i i thc foimi
of a profile cluriiiq the clinical inter-\-icw. Cliwtei.
objectivity in soniatot~-piiig s obtained byt a l k of l wi~c,
nusclc irth tilid subcutaneous fat measure-ments for each licigh t
mid pondci-al i d e s group. These
tahlcs provide a siniplc and reliable giiidc to doiiiinaiice
in
-
7/28/2019 Parnell 54
26/32
234 It. W. PARSELLall but a small percentage of individuals who
show reversalof dominance in trunk and limbs.2. agreement to half a
unit was obtained in 90% of allsomatotype ratings between estimates
by deviation chart andtables and estimates made by the long
photometric method.Between estimates by deviation chart and tables
and photo-scopic estimates by espcrt somatotypists agreement to
halfa unit was found in 87.3%.3. The prospect for somatotyping by
deviation chart andtables in older meii and woineii subjects is
discussed.
LITFRATtTRE CJTEI)EDTT-AKDS,. A . W. 19,it) Oliwrmtions on tiit.
djstriliiition of subciitniwoosH ~ s T . 2. E. 1952 Hum:iii
ronstitiitiou: a n appr:iisul. Am. J. Php. Bntbrop.,SHELDOS,.H.
1952 Private communication.SKELDOS, . H., E. M. H A R T L AND E.
MCnWlroTT The varieties ofSHELDOS,. H., S. S. STEVENSND W. E.
TUCKER1940 The varieties ofTANNER, . M. 1951 Current advances in
the study of physique. Photogram-
fat. Clin. Sci.. 9: 2.79.10 : 55-73.
1949delinquent poutli. Harper and Brothers, Ncw York.human
physique. Harper and Brothers, New York and London.metric
antliropometry and an sndrogpny scale. Laneet, i, 571.1952 The
physique of students. Lancet, ii, 405.
T%KSER.. N., A4ND J. S. WEIKER1949 The reliability of the
photogrammetricmethod of antliropometrg, with a description of a
miniature cameratecliniqnc. Bin. J. Phys . A4ntIirop., 7 :
145-186.
-
7/28/2019 Parnell 54
27/32
SOMATOTTPISG RP -4STHHOPOMETBT 233
Humerus , cmFemur , em
-
7/28/2019 Parnell 54
28/32
236 R. W. PBRSELL
Mew aged 17-24Height/ d w e i g l t 12.50-12.95 inclusive
E E I G H TIXCHES E S D O - M E B OESTIMATE
Humerus, em 6.0 6.2 6.3 6.5 6.7 6.8 7.0 7.2 7.4Femur, ciii 8.6
8.8 9.0 9.2 9.4 9.6 9.8 10.0 10.2
-
7/28/2019 Parnell 54
29/32
SOMATOTYPING BY ANTHROIOMETRY 237Men aged 17-94
Height/ V weight 13.00-13.45 inclusive~
2 ZP 3 31 4 48 5 5% 6EIGHT RFDO-YES0INCHES ESTIMATE_ _Humerus,
cm 6.1 6.3 6.4 6.6 6.8 6.9 7.1 7.3 7.4Femur, cm 8.5 8.7 8.9 9.1 9.3
9.5 9.7 9.9 10.1
-
7/28/2019 Parnell 54
30/32
338 R. W. PARRELLM e n aged 17-14
Height/?- 13.50-13.95 inclusive....
1 14 2 21 3 ::g 4 41EIGHT ENDO-ME80ISCHES BRTIMATE 5-Humerus, cm
5.9 6.1 6.2 6.4 6.6 6.7 6.9 7.1 7.2Femur, cni 8.2 8.4 8.6 8.8 9.0
9.2 9.4 9.6 9.8
-
7/28/2019 Parnell 54
31/32
SORIATOTYPIXO B T ASTHROPOMIETRT 239
H E I G H TINCHES
Yen aged 17-34Height/ QGizbt 14.00 + _ _ - .~
$ 1 If 2 z t 3 3*NW-MESOESTIMATE
< 0Humerus , em 6.2 6.3 6.5 6.7 6.8 7.0 7.2Femur, em 8.7 8.9
9.1 9.3 9.5 9.7 9.9Biceps, cm 22.6 23.7 24.8 25.9 27.0 28.1 29.3Cal
f , em 28.4 29.5 30.5 31.5 32.5 33.5 34.6T o t a l f a t , mu1 16
19 23 28 34 40 49
i 0 - i l . 9
i 2 f
Humerus . cmFemur , cmBiceps, cmCal f , cmTo t n l f n t ,
mm
~~ .. _ _ _Hu me r u s , cmF e mu r , emBiceps, cmCal f , C l l
l
6.38.9
22.629.516
6.59.1
23.130.2
6.5 6.7 6.89.1 9.3 9.5
23.7 24.8 25.930.5 31.5 32.519 23 28~- -_--6.7 6.8 7.09.3 9.5
9.7
24.2 25.3 26.431.2 32.3 33.3
7.0 7.1 7.39.7 9.9 10.1
27.0 28.1 29.333.5 34.5 35.634 40 49
_ _ _ _ ~~~~~7.2 7.3 7.59.9 10.1 10.3
27.5 28.7 29.834.3 35.3 36.3
T o t a l f a t , nim 16 19 23 28 34 40 49
hIORFOLOGIA IKFASTII, (CRECIMIENTO) . Juan Comas. (Reprintedfrom
Paidologia by Jose Peinado Altable, Chapter VI, pp. 221-349.136
pp.. 17 graphs aiid figures, 89 tables. Mexico, 1952.) -A sum-m a r
- account of child growth which includes tabular material fromthe
world literature on the subject. Of special interest is the
coverageof the Latin American data and literature.
-
7/28/2019 Parnell 54
32/32
240 AMERICAN JOURNAL O F PHYSICAL A S T H R O P O L O G T
E A S I C B O D Y A ~ E A SCRENE S T S O F &IiOOL &FS C
H I L D R K S . fisEdgar Martin. U. S. Department of Health,
Education, and Welfare,Washington. 74 pp. 1953.- This free booklet
published by the U. S.Department of Health, Education, and Welfare
is straight appliedphysical anthropology and provides normative
data on a variety ofbody measurements under the assinnption that
they will assist inthe proper manufacture of school equipment.
Although soine anthro-pologists inay object to weighted means and
common standard devia-tions obtained by combining divrrse series
ineasiired o w r a 20-yearrange, it is not likely that todays data
would be too very different.And while the definition of nasioii as
a point between the eyesinay be inadequate for routine
anthropometry, the imprecision shouldnot affect nianufactnrers of
blackboards, wintlows, or child-size wash-basins. Fortunately, all
iiiensnrenients arc rrported in inches aiidtenths of an inch.
B i r . Martin carefully distinguishes brtween the static
ineasure-ments customarily taken and the dpamic measurements that
usuallycaiinot be extrapolated froni thciii . IVhile it iq
important for iiianii-facturers to realize these limitations of
conwntional body measure-ments, it is even inore important f o r
them t o realize that the arith-metic inem is iisnally the p o re s
t possiblr statistic to iise when hiinancomfort is concerned. Wh at
is not included in this booltlet, naiiiclpthe theory of applied
physical anthropology. is inissiiig breansc ithas not been written
by applied physical anthropologists thrmsc~l~-rs.-S. 11. G A R S
.
MAN I?; EToLwrrnN. 11. R . Sahni. ( 2 7 2 pp., 51 x 9.
illnstratcdby Ihmini Sahni. 13 shillings. Longiiians Grern and Co..
London.1933.)- his account of human evolution by an Indian author
isintended maiiily for Indian readers, and incliidrs a siirre- of
whatis known of the prehistory of Jndia and the neighboring
coiintries.- G . W. LAKKER.
