close

Вход

Забыли?

вход по аккаунту

?

Blood groups of American Indians.

код для вставкиСкачать
BLOOD GROUPS O F AMERICAN INDIANS
WILLIAM C. BOYD
Evans Memorial, Massaclwetts Memorial Eospitaln, and Boston Unicereity
School of Medicine, Mlassachuaetts
ONE FIQUBE
I n the books and papers touching on the subject of blood
groups in relation to anthropology, the statement is still
commonly found that pure-blood American Indians belong
entirely to blood group 0.l In some of the more recent
articles an exception is made in the case of the Bloods and
Blackfeet, where Matson and others (’33, ’36, ’38) found
percentages of A as high as 83%. This does not seem to
have weakened the impression, however, that all the other
tribes belonged entirely to group 0 before 1492, and that
such amounts of A and B as have been reported are due to
white mixture. Even the pioneer Hirszfeld says in his latest
book (’38)“Dam l’existance des races dkpourvues des propri6t6s A et B, comme les Indiens, les savants voient la preuve
qu’aux temps pr6historiqucs les races humaines primitives
appartenaient au groupe 0 et que A et B seraient une acquisition ulthieure.”
It seems not to have occurred to any of the earlier workers,
even after the Bernstein (’25) hypothesis of blood group
heredity was established, to calculate mathematically the
chances that the observed Indian results could really be due to
race mixture. I n 1935 Wyman and I (’35)pointed out that in
some of the earlier work on Indians, it had been observed
that there was rather more A, proportional to the amount of
B found, than could readily be accounted for on the basis of
Candela ( ’39) has, however, recently called attention to the increaeing need for
recognition of the presence of A in North American Indians in general.
215
AMERICAN JOURNAL OF PHYSICAL ANTAROWI~OOY, VOL. XXV, NO. 2
JULY-SBPTBYBRB, 1939
216
WILLIAM C. BOYD
mixture with the white man. This statement was not supported by any detailed calculations, however, and seems to
have made little impression.
I n the present communication, I propose to give a summary
of all the results to date on the blood groups of the American
Indians. No existing summary is now at all complete or
critical. I n addition, I propose to test each result mathematically to determine the probability with which it could be ascribed to the result of mixture of a pure race of group 0 with
a race having the characteristic European distribution.
METHODS
Knowing the genetic mechanism involved, it is comparatively simple to calculate the resulting composition of a
mixed race, if we know the proportion in which the two parent
races were mixed. It is only necessary to take a weighted
average of the gene frequencies of the two races.
I n the present instance, we have, by hypothesis, race I is
Race
entirely group 0. Therefore p =0, q =0, r =
I1 is to be European in blood group distribution. The various
European races which early had opportunity to mix with
the Indians differ, but only slightly, from each other in this
respect. Since me might suppose that the undetected mixture in the Indians took place some time ago, it seemed appropriate to take for race I1 predominantly the English blood
group distribution. This is also fairer to those who believe
that the A and B in most North American Indians is due to
European mixture than taking, for example Snyder's results
for 20,000 white Americans, where rather more B was found.
The results of four determinations of blood group frequencies
in England, compared with a typical French, and a typical
Spanish distribution, are given in table 1.
'p, q, and r represent the frequencies in the population of the genea for A ,
B, and 0, respectively. According to the well-established Bernstein theory, they
may be calculated by the formulas, p = V F O - V O i q=VB+O
- V q
r = V 3 , where the large letters repreaent the frequencies (%/loo) of the blood
groups.
217
BLOOD GROUPS OF AMERICAN INDIANS
Since it did not seem too unreasonable to assume that these
three ethnic groups had contributed in roughly this proportion to the production of racial hybrids in North America, an
average was made of all the gene frequencies tabulated in
table 1 which, allowing for a slight adjustment to make the
sum exactly 1.00, gave p =0.281, q =0.059, r =0.660. This
is equivalent to a racial group having blood group frequencies
of 0 =43.5%, A =45.0, B =8.2, AB =3.3. Examination
of other possible distributions will show, however, that the
net result of the calculations we are about to undertake would
TABLE 1
Results of four English, one French and one Spanish, blood group studies
AuLhor
Numbrr
tested
Taylor and Prior 422
Kirwm-Taylor 500
Hirszfeld and
403
Hirszf eld
Penrose and
Penrose
1000
Dujarric and
Kossovitch
1265
Bote-Garcia
296
Per cent of proup
A
B
Qcns f r e q u c d e s
AB
P
P
r
47.9
42.4
8.3
1.4
0.250
0.050
0.692
40.4
46.8
9.6
3.2
0.297
0.072
0.637
44.6
44.6
7.7
3.0
0.277
0.055
0.668
43.2
47.7
6.4
2.7
0.295
0.048
0.655
39.8
42.3
11.8
6.1
0.276
0.088
0.632
44.3
46.6
5.4
3.7
0.288
0.040
0.666
0
not be greatly altered by the choice of other possible, different,
European distributions.
From the above values of p, q, and r for the hypothetical
races I and 11, we can readily calculate the blood group gene
frequencies of various hypothetical mixed races, arising from
the mixture of 0, a, +,4, etc., white blood with pure Indian
group 0 blood. From these we get the blood group frequencies
themselves by the equations 0 =r2, A =p2 2pr, B =q2
2qr, AB =2pq. A graph of the resulting calculated relation
between the group frequencies of the mixed race and the
percentage of European blood is shown in figure 1.
For Mexico and South America it seemed on the whole
fairer to make the calculations on the basis of Spanish mixture
+
+
218
WILLIAM C. BOYD
alone, and a similar table of frequencies for the mixed race
was prepared using the Spanish figures from table 1.
The actual amount of European mixture in the various
groups of Indians examined for blood group frequencies is
100
90
00
10
60
50
40
30
20
10
Figure 1
hard to ascertain. In some cases, as in the Haskell Institute,
there is doubtless a good deal of European admixture. I n
some cases the authors tell us that their Indians were 'pure';
BLOOD GROUPS OF AMERICAN INDIANS
219
more generally they tell us nothing. So the device was
adopted of calculating for each group of Indians the assumed
percentage of European mixture which produced ‘mixed’
blood group frequencies giving optimal agreement with those
observed. This would seem to be all the proponents of the
‘mixture’ theory could ask.
I n order to find the amount of mixture giving optimal agreement, it seemed sufficient to find that percentage of European
blood for which the calculated mixed frequencies gave together with the observed frequencies, the smallest value for
xz (compare Strandskov, ’31).
It might be possible to derive a general expression connecting the observed frequencies with the assumed percentage of European mixture which will give a minimal value of
x2, but the algebraic difficulties seemed so great that instead,
for each item, values of x2 for several points in the neighborhood of the minimum were calculated from values of the
‘mixed’ frequencies corresponding to various assumed values
of white mixture, the values graphed, and the minimum determined from the graph. This method undoubtedly yielded
results of ample accuracy for our purpose, and in fact it
could often be seen at the outset that the minimal value of xz
would not be significant. I n such cases often only an approximate value was obtained, designated by c in the table. From
the minimal value of x2, the probability that the difference between the corresponding ‘mixed’ blood group frequencies, and
those actually observed, could be due to chance alone can be
determined from standard tables (Fisher and Yates, ’35;
Pearson, ’30),assuming two degrees of freedom. These probabilities are given in the column Pa (= optimal probability) in
table 2.
It was shown by Bernstein (’30) that the probability that
an observed blood group distribution could be a random
sample of a population in equilibrium, obeying the laws of
the Bernstein hypothesis of blood group inheritance, can in
each case be calculated. These probabilities are given in
table 2 under P, (= genetic probability). It is of course
clear that we have no right to demand that Pabe of a reasonAYERICAN JOURNAL OF PHYSICAL ANTHBOPOLO(IV, V O L XXV, NO. 2
Blackfeet
Agency,
Mont.
Browning
Mont.
Alberta,
Canada
Browning
Mont.
Cardston,
Alberta
Cherokee
Agency,
N. C.
Lake
Atlia basea
Lake
Athabasca,
etc.
Edmonton,
Alberta
Mackenzie
Riv. basin
Montana
Montana 1
W Montana
Blackfeet p
Blackfeet p
Bloods p
6
7
Bloods
Bloods
Cherokee I
Chipewyan m
cm?P
Cree p
Dogrib
Flatheads p
Flatheads m
Flatheads
(Salish) m
Hopi
E
l3
0
9
9s
10
11
12
13
14
15
16
17
yon, Arie.
H.Agency,
Keama Can-
Montana
Blackfeet rn
Blackfeet
Blackfeet p
Grant ( '36)
Beaver p
68
77
Matson ( '38)
Grant ( '38)
52
53
691
258
45
Qrrnt ('33)
250
105
69
24
84.6
77.4
53.8
51.5
63.6
73.5
89
80
74.4
19.1
17.4
16.7
24.3
103
45.5
23.5
15.5
22.4
69
Matson ( '38)
Matson ('38)
Ma tson &
Scbradrr ( '33)
Snyder ('26)
TABLE 2
107
.
235
115
123
Grant ('30, '30)
Snyder ('26)
N~mbsT
15.4
17.0
39.5
42.2
36.4
25.0
11
10
16.0
80.0
81.2
83.3
74.8
76.7
50.6
76.5
83.7
0
4.7
1.9
4.8
0
1.5
0
10
0
3.7
1.9
1.G
0
0
0
0
2.4
0.9
0
7.2
1.4s
0
1.0
0.9a
1.8
0
0.8
0
AB
0
0
0
0
2.1
0
0
t68tEd
0
A
B
North America (except Mexico)
40
52.5
47.5
0
Matson &
Schrader ( '33)
Matson &
b h r a d e r ( '33)
Levine.Matson
6r Schiader
('35)
Mntson.Levine
ds Sehrader
( '36 1
Matson &
Schrader ( '33)
Levine,Matson&
Sehrader ( '35)
Matson ('38)
Matson ('38)
Invsatigalorr
Tram
Pl4C6
Peace River,
Canada
Gleichen,
Alberta
Montana
Item
1.00
0.76
1.00
e18
1.oo
c25
83
56
78
64
44
39
50
100
100
100
100
100
100
100
100
88
0.69
0.0008
>0.10
>0.10
1.00
>O.lO
>0.10
0.42
1.00
>O.lO
1 .oo
A
B
A
c0.41
0.17
0.00
0.014
0.044
0.40
c0.59
0.0058
0.00035
<lo-'
<lo-"
0.00047
<lo-'
<lo-'
0.00087
<lo-#
<lo-'
0.105
Kwakiutl (except Bella
Bella & Kitamaat)
Loucheux
Micmac
Muskhogenn I
20
22
23
Navajo
Navajo p
Narnjo 111
Navajo
Nuotka
Piegall
25
26
2i
28
29
30
Muskhogem I
Kwakiutl
(Kitamaat)
19
21
Kwakiutl
18
8.6
30.6
91.4
69.1
28
279
137
81
622
457
Gates ('38)
SnFtler ('36)
Landstoiner 6;
Lcrine ( '29)
Mackenzie
Riv. basin
Nova Seotia
Clloctnw( !Iiik:rsnw
S:i nn tori nni,
Talihiirn,
Okla.
(%octnwCtikasuw
Sanatorium,
Tolihina,
Okla.
N. M.
Ft. Definnce
Broeket,
Alberta
B. c.
Michets,
Ariz.
Ft. Dcfiarice
& st.
Miclinrls,
Ark
?
('onst,
& St.
Matsoii ( '38)
& Puruhnta
('35)
( '34) 8; Ride
42
19.1
99.1
98.4
G1
112
Rife ('33)
Gates & Uarby
56.7
00
Nigg ('26)
72.9
62.0
89.7
100
85.4
123
const,
13. C.
80.9
0.9
1.G
41.0
26.9
0
0
0
1.7
0.2
0
Oe
0
0
0.2
0
0
0
0.2
1.1
0
0
Oe
Oe
08
4.0
0.7
0
0
32.9
9.4
2.4
1.7
0
12.2
39.7
58.6
58
7.4
Gates & Darby
( '34) & Ride
& Furuhata
('35)
Gates & Darbv
( '34) & Rid;
8; Furuhata
( '35)
Grant ('38)
92.6
94
Gates & Darby
('34)
Bells Bells
Br.
Columbia
Kitameat,
Rr.
Columbia
100
1.00
1.00
>O.lO
>O.lO
>0.10
100
c3
e3
73
47
5s
c13
A
f
1
A
A
A
A
0.000013
c0.93
c0.80
0.15
0.00013
<lo"
c0.43
0.36
e0.60
A
A
63
el5
0.55
>O.lO
1.00
1.00
c0.95
f
0
0.17
A
0
c23
>O.lO
00.48
A
1.00
70
c13
>Oslo
1.00
N. M.
S & F Sans-
Pueblo, Towa
Pueblo, v
Sac & Fox
(Algonkiaii)
34
35
36
8. Ageiicy,
Ft Washakie, Wyo.
Crow Creek,
Agency, Ft.
Thompson,
6. D.
Ft. Totten,
N. D.
Ponca
Whiteagle,
Okla.
Canton, 6. D.
Rosebud
Slioahuiie
Sioux?
Sioux p
Siouan
Sioux?
Sioux
39
40
41
42
43
torium,
Toledo, Ia.
Coast,
B. C.
38
Saliah
Taos, etc.,
N. M.
Pueblo,Tiwa
33
Jeniez, etc.,
N. M.
San Juan,
etc., N. M.
Pueblo, Tewa
g 37
LO
Place
Cochiti ,
etc., N. M.
33
TrQs
Pueblo,
Keresan
Item
31
310
Alleii e t al.
( '35, '37)
Alleii &
Schaeffer
95
40
84.0
100
Snyder ( '26)
Snyder ('26)
89.6
48
Matson &
Schrader
( '33)
Snyder ('26)
70.5
70.0
25.2
27.5
11.0
10.4
7.0
91.0
20
26.6
13.3
20.0
19.2
9.9
45.0
100
A
11.7
51.6
Snyder ('26)
80
71.6
84.5
78.4
75.4
90.1
60
50
60
0
88.0
Snyder ( '26)
('35)
Gates & Darby
( '34) & Ride
& Furuhata
Snyder ('26)
( '35)
('35)
307
203
Allen &
Schaeff er
('35)
142
353
t88tad
Allen &
Ekhaeffer
( '35)
Allen &
Schaeffer
InV88tigdOr8
TABLE a 4 o n t i n u c d
Number
B
AB
p,
0
0
0.37
0.48
49
48
A
At
(B)
54
6 X lo-"
2.0
3.O
0.67
0.65
0.01
c.0.58
A
e18
1.00
0
0
c0.87
>O.lO
0
B
0.35
0.42
e15
A
A
c0.18
0.073
A
A
0.51
c0.26
A
A
p.
c0.39
0.19
84
47
c25
85
40
c15
A
Ex
A
2.0
4.2
2.5
we
c18
85
1.00
>O.lO
>O.lO
>O.lO
>O.lO
1.00
>O.lO
>O.lO
Oe
0
0.6e
Oe
0
0
0.3
1.6
1.6
0
1.6
1.6
1.6
5.4
0
0
N
N
Klenltu,
B. C.
Ft. Bidwell,
Calif.
Haskell
Inst. Lawrence, Kan.
Haskell
Inst. Lawrence, Kan.
Haskell
Inst. Lawrence, Kan.
Haskell
Inst. Lawrence, Ean.
Kan., N. M.,
& Okla.
Cliemawa,
Ore.
Mouth
Mackenzie
River
Peace Riv.,
Canada
School, Hayward, Wis.
Stoney p
Taimshian
Tsimshian &
Kwakiutl
1
V
47
48
49
51
V
V
V
vm
v Athabascans
V
53
54
55
56
57
58
59
V
52
I0
Coast,
B. C.
Stoney
46
Alberta,
Canada
Morley,
Alberta
Slave
45
Mouth
Mackeneie
River,
Canada
Mackenzie
Riv. basin
Slave
(Athabasean)
44
77.7
54.7
82
124
316
862
64
33
61
40
Landsteiner &
Levine ( '29)
Nigg ('26)
Coca & Deibert
( '23)
Matson &
Schrader ( '33)
Gates ( '29)
Grant ( '36)
Snyder ( '26)
75.0
62
70.9
66.1
120
Downs, Jones &
Kwrber ( '29)
65.0
66
112
( '23)
Coca & Deibert
72.2
89.1
54
73.5
46
49
Matson ('38)
67.1
63.4
158
Grant ('38)
36.5
41
41
Qrant ( '38)
87
Gates & Darby
( '34) & Ride
&hmhata
('35)
Gates & Darbv
('34)
Snyder ( '26)
38
Gates ('29)
17.5
38
18
32.8
20.2
27.2
33.9
35.0
27
8.7
34.2
27.8
26.5
32.3
63.5
13
5.0
0
0
9.4
2.1
1.6
0
0
6
2.2
2.4
0
0
0
0
0
2.5
0
0
3.1
0
0.3
0
Ob
Id
0
Oa
Oe
0
0.6a
Oa
0
0.25
1.00
1.00
0.23
0.37
>O.lO
1.00
1.00
0.32
>0.10
>0.10
1.00
1.00
1.00
1.00
1.00
49
66
c28
80
36
48
68
60
55
el8
62
48
46
57
100
c20
B
A
A
B
A
A
A
A
0.26
0.073
c0.51
0.53
0.0068
0.019
0.0093
0.009
0.92
c0.95
BI
1
0.46
0.20
0.28
A
A
A
0.0076
0.016
A
A
c0.59
A
I&
fs
N
Argentina
Argentina
Paraguay
Near
Temue0,
Chile
Caraja
Gliorote
Chulpie
Colla
Guarani
Mapuche (=
Araueano )
67
68
69
70
71
'Am. Indians'
64
66
Maya?
63
Gran Boree
& Green
Cassava
Falls,
Surinam
Araguaya R.,
Brazil
Argentina
Maya
62
Alkuyana
No. Mexico
Maya P
61
65
Dzitas,
Yucatan
G%icliBn Itz5
& Chankom
Upper Ushniacintla
val. & Yucatan Peiin.
Yucatan
Placs
Maya m
Tribe
60
Item
63.5
100
75.6
107
382
89.1
115
147
Mazza, Schiirmanii & Gutdeutsch ( '36)
Mazza et al.
( '36 )
Mazza e t a].
( '36 1
Ribero ('36)
Rahm ('31)
39
100
100
61
Golden ('30)
28.1
16.7
1.6
1.3
9.9
A
0
B
0
46
0
0.6
0
6.2
0
17.2
1.00
0.62
A
64
1.00
0
0
36.5
0
0
1.00
0
0
A
B
0
B
B
Al
(Bl
B
Ex
0
100
0
78
44
c3
15
67
w e
c18
0.69
1.00
0.07
0.23
1.00
>O.lO
0.16
p,
1.00
5
0
0.9
1.4t
0
0.5
0
AB
0
51
0
11.8
5.4
0
0.5
5.0
B
0
10.9
5
0
Kouth America
59.2
76.5
98.4
97.7
85.1
Mexico
55
70
338
Kahn ('36)
738
124
Rife ('32)
nedy ('29)
Moss & Kennedy ( '29)
223
Goodlier ( '30)
Moss & Ken-
202
lerted
0
TABLE 2 4 o n l i n u r d
Number
Goodner ( '30)
Invssti~dorr
1.00
0.00017
0.038
1.00
~0.30
<lo-#
1.00
<IF
<lo-.
c0.80
0.0000ls
0.0064
Po
01
tu
N
Na.v:iriiio Is.,
Tierra
tlel Fiiegu
No. Prov.,
Argentiiin
Chile
Peru
Peru
Toba
Yagliari
V
(White?)
ni
P
75
56
77
79
80
33
194
46
18
135
Mnzza 6; Franke
120
( ’27)
242
Meza ( ’33)
Arce Larrita
1372
(’33)
200
Arce Larreta
( ’33)
Moss & Ken187
nedy (’20)
Mazm et 01.
(’361
Muz& ct nl.
( ’36)
R&ii~ ( ’31)
hfazza ct Ill.
( ’36)
Ralrm ( ’32)
55.6
100
52.9
75.1
100
9.0
98.5
82.6
94.4
96.3
0
0
13.4
32.6
7.2
12.8
14.1
25.1
0
91.0
0
0
0
0
0
0
1.5
17.4
5.6
3.7
5.9
0
1.7
3.G
0
0
0
0
0
0
050
0.0135
100
0
100
0.23
<ioJ
1.00
0
100
c3
c28
C8
c5
1.00
1.00
1.00
1.00
1.00
1.00
B
0
B
B
0
B
Al
A
A
A
<lo-.
1.00
<l&’
<<1P
1.00
<lo”
c0.83
c0.53
c0.93
c0.63
E, is the ‘element in excess’ in tlie observed, a8 opposed to the cakulated figures. P. is the ‘optimal probability,’ or the chance that the
difference between the observed and calculated frequencies is due to random sampling. p s i g d l e s pure, or ‘said to be pure,’ tn denotes mixed,
~1 signifies various, a significs that the per cents given were calculated by me from the original data, b that I have recalculated the per cents,
d signifies that the series may inelude, or be included in, another series in the table, e that I have combined, by addition or subtraction, two
series originally reported separately. The letter o before a figure signifies that it is only approximate. Reaulta in italica are results cornmented on in the text as being improbable.
+
Explanation of symbols used in table 2. Resulta are arranged, under each main heading, alphabetically according to tribes. The symbols
0,A, B, and AB refer to tlie percentage of these groups found in the population. P, represents the probability that the observed frequencies
could represent a random sample from a populatioii having some values of p -f- q r = 1 (i.e., genetically obeying Bernstein’s theory). W,
represents the “optimal per cent of white mixture,” calculated 88 described in the text t o give values agreeing o p t h l l y with the observed.
81
Huancabaniba dc
Perico,
Peru
Argciitiiia
Pilnga
74
ni
Tierra del
Fucgo
Argeiitiriu
Ona
73
i8
Argentina
Matuco
72
226
w u m r c. BOYD
able magnitude if P, does not also fulfill this condition, since
we have made the (well substantiated) Bernstein theory the
basis of our calculations. It will be seen that with few exceptions, P, indicates that the data are compatible with the
Bernstein theory. The exceptions are doubtless due to lack
of genetic equilibrium, or errors in the determinations, or both.
I n table 2 is given a summary of all blood gronping data
known to me on American Indians, where more than a few
persons were tested. After the blood group percentages,
there is also given P, (see above), and W, (=optimal percentage white blood) is that percentage of mixture which
must be assumed to make x2 a minimum. Then, in order,
comes E, (= element in excess), the blood group factor more
plentiful in the observed than in the calculated results, and
Po(see above).
Naturally it is found that when either factor A or B is
completely absent from a population, P, is 1.00, for any such
population could be explained on the basis of some proportions
of the other two genes. When the percentage of A or B was
extremely small, I have not calculated P,, as even there the
test loses its value.
If the observed results show an excess of AB, this is signified by writing ( B ) under the heading Ex.
DISCUSSION
Perhaps the most salient feature of the table is the extreme
rarity of results showing the presence of group 0 only. Of the
eighty-two items, there are only six of which this can be
said. This is notwithstanding the fact that many of the
groups tested were labeled ‘pure’ by the autbors.
If we next glance at the column headed Po,
we find that in
thirty-three, or 40%, of the instances the value of the probability is less than 0.05, so that, if we follow the usual statistical
convention, there is a statistically significant difference between the observed blood group frequencies, and any possible
frequencies which could result from mixture of a pure group
0 stock with the white man. This in spite of the fact that
BLOOD QROUPS OF A M E R I C A N I N D I A N S
227
most of the numbers examined were very small so that a
large difference is required to give a significantly low value
of Po. I n most of the thirty-three cases the probability is
enormously less than 0.05, often so low that it could not be
found in the tables examined. We may regard it as certain
in such instances that these Indians do not owe their blood
group picture exclusively to mixture with the Europeans.
The four cases were P, is also less than 0.05 should, strictly,
be left out of account, as mentioned above.
Several of the thirty-three items concern the Bloods and
Blackfeet, mentioned above, and taken cognizance of by most
modern authors, but definitely low probabilities are also encountered in the case of items 10 (Chipewyan), 14 (Flatheads),
25 and 26 (Navajo), 30 (Piegan), 45 (Slave), 46 (Stoney),
52-54 (Indians at the Haskell Institute), 55 (Coca and
Deibert's pioneer results), 60, 61, 63, and 64 (Mexico), 66, 71,
76, 78 (South American tribes).
Thus it is not only the Bloods and Blackfeet, and the
notorious Yaghans, which cannot be explained by the hypothesis of mixture, but a fair number of other tribes and
mixed groups. Some of these results have been in the literature for some time, for example item 26, and items 55, 52,
53, and 54. The earlier workers, of course, did not know the
great variation in group frequencies to be found in the world,
and inertia has probably prevented later workers from reexamining the data in the light of that knowledge; to both
groups the hypothesis of mixture has seemed at least until
recently the most obvious and satisfying explanation.
But if we examine the column headed Wo (optimal per cent
white), even more surprising results can be noted. Recall
that we have, by our method of calculation, allowed the proponents of the mixture theory the most favorable frequencies
out of the whole set which could result from mixture with
Europeans. (Even so there were in many cases significant
discrepancies remaining.) We did not ask if the percentages
of assumed mixture were of such magnitude as to be reason-
228
WIILIAM C. BOYD
able. If we now examine the values in column W,, I think
it can be seen that in a large number of cases they are not.
It is hard to estimate the percentage of European blood
in present day Indians, but it may be doubted if on the whole
it exceeds 25%. Occasional individuals with this much, or
more, European blood, may have been included in the results
here summarized, but it hardly seems reasonable that whole
groups of Indians can possess this much European blood.
Certainly we may suppose that Indians which were half
white would differ significantly in a number of respects from
pure Indians.
I n the column W,, we find that in fifty-three instances, or
65% of the total, the percentage of European blood demanded
exceeds 25, and in thirty-three instances (40%) it exceeds 50.
In many instances even the ridiculous assumption of 100%
European blood does not suffice to give frequencies agreeing
with those actually found. This is to be expected in tribes
such as the Bloods where the percentage of A is higher than
in any white race, or the Yaghans, where the percentage of
B is greater.
Even in the case of less well-known examples, the discrepancy is still extreme. The hypothesis of mixture may be
said to explain the blood group picture of the Kwakiutl in Kitamaat, in the sense that certain frequencies which could result
from mixture do not differ improbably from those found, but
it is necessary to postulate that these Indians are 70% white.
Any attempt to get along with much less white blood will
result in frequencies giving too low values of Po. I n the
case of the Navajo (items 25 and 26) the optimal percentage
of European mixture is 47 to 53, but even this does not give
reasonably high values of Po. I n the case of the Shoshone
Indians, tested long ago by Snyder, it is necessary to suppose 85% European blood. Similar results are encountered
in other cases.
Another significant fact emerges from an examination of
the column headed E, (element in excess). I n North America,
in each case, with perhaps twelve exceptions, the blood
BLOOD GROUPS OF AMERICAN INDIARS
229
group element A is found to be in excess of the value calculated from the hypothesis of mixture. Clearly there is more
A in North America than was dreamed of by the proponents
of the mixture hypothesis.
The results for Mexico and South America are in contrast
to the above. The results for Mexico show high percentages
of 0, but are puzzling, and in four out of five items the element B seems to be in excess of the optimal values calculated
from the assumption of Spanish admixture. In South America
there are four instances (items 66, 71, 76, and 78) where the
element B is considerably in excess, and Pois very small in
all these cases. I n certain other cases A is in excess, but in
only one of these can we be reasonably sure the A found could
not be due to Spanish mixture. And, at last, we find some
tribes which very probably possessed, before white contact,
neither A nor B.
The above findings are cumulative in their effect, and it
would seem that we may say without hesitation that many
tribes of American Indians had significant amounts of blood
factors other than 0 before the arrival of the white man.
N70rkers who reject this conclusion will probably support
their position by two main arguments : inbreeding, and mutation. It does not seem to me that either or both are sufficient.
Inbreeding, in particular, may be easily dismissed. Hogben
('32-'33) has derived formulas to express the effect of inbreeding on the distribution of characteristics in a population. From them we may calculate that if we start with a
population such as thc Indians at the Haskell Institute (item
54), and continue, for a very long period of time, to mate
first cousin to first cousin (about as high a degree of inbreeding as can have obtained in any Indian tribe), the ultimate
result is a slight increase in the percentage of group 0;
from 70.9 to about 71.1%! Therefore we cannot expect inbreeding to change markedly the frequencies of blood groups
in an isolated group.
It may be suspected that some authors who use the term
inbreeding do not really mean it in the same sense as the
230
WILLIAM C. BOYD
geneticist does, but have in mind the picture of the offspring
of, say, one white man with one or two Indian women,
multiplying in isolation to produce a small tribe which might
have one of the blood group distributions observed. Some
workers will attempt, at least in conversation, to account for
the blood groups of the Yaghans in a similar way. We cannot
certainly say that this has never happened, but it may be
doubted if the situation imagined is at all common. Also
there are other objections; not many Indian tribes have been
on the increase during the last few centuries; a tribe with so
much white blood ought to show other non-Indian characteristics ;it would be hard to explain why nearly all the European
men who chose to misbehave thus in North America belonged
to group A.
The hypothesis of mutation, to explain the blood groups
of the Indians, is still popular with some authors. The idea,
apparently once naively held, that a single mutation could
explain the world distribution of A, has given way to the
idea of an increased rate of mutation in certain areas. But
a thorough analysis of the data suggests that in all probability North (and even South) America was fairly riddled with
A (and B). So instead of an ‘island’ of mutation in the New
World (Gates, ’36), we shall need a whole archipelago. I
have previously discussed at length the mutation hypothesis
(Wyman and Boyd, ’35, ’37), and pointed out that the assumption of the necessary number of independent centers of
mutation involves the assumption of a staggering coincidence.
When it is once seen that the Indians (at least some of them)
, and B, and we know that there is
originally possessed 4
plenty of A and B in Asia for that in America to have come
from, it seems definitely a wasteful multiplication of hy, and B in America is of indepotheses to suppose that the 4
pendent origin.
BLOOD GROUPS O F AMERICAN INDIANS
231
CONCLUSIONB
It will be convenient to discuss the conclusions separately
for each of the three geographical divisions.
North Amerka. On the basis of the inadequate data at ourc
disposal, and the above analysis of them, we may provisionally
classify North American Indians into four classes: I) those
undoubtedly possessing considerable 3,in some cases over
50%, before the arrival of the white man, 11) those possessing, or probably possessing, some A, 111)those probably pure
group 0, IT)those impossible to classify on the basis of
the available evidence. This classification follows :
I (much A) Beaver, Bloods, Blackfeet, Navajo, Piegan,
Shoshone, Slave, Stoney, Tsimshian.
I1 (some A) Cree, Dogrib, Flatheads, Hopi, Kwakiutl,
Pueblo (Keresan, Tewa, Tiwa, Towa), Sac and Fox,
Salish, and Indians at the Haskell Institute.
I11 (pure 0) Loucheux, Muskhogean (items 23 and 24),
Nootka.
I V (unclassifiable) Cherokee, Chipewyan, Micmac, Sioux.
It must be left to authorities on North American tribal
relationships to decide if tribes in class I1 could be the result
of mixture of tribes of classes I and III.
Mercico. The data from Mexico are puzzling, and particularly inadequate. Pure Mayas may have been pure 0. Could
the apparent excess of B in four of the Mexican results be
due to non-Spanish (Negro?) mixture?
South America. We have less data for South than for North
America, and the situation would seem to be even more complex. Perhaps we might tentatively classify the South American Indians under the same categories as North Americans,
adding a fifth class (=V), Indians possessing B in amounts
in excess of those reasonably attributable to European
(Spanish) mixture, plus possibly some A. There seem to be
no examples of class I in the South American data at present
available. Thus we have:
232
WILLIAM C. BOYD
I1 (some A) Colla and (perhaps) Pilaga.
I11 (pure 0) Alkuyana, Chulpie, Guarani, Mataco, Ona,
Toba, Indians in Argentina (item 77), and in Peru
(item 80).
I V (unclassifiable) Chorote and (perhaps) Pilaga.
V (much B ) Caraja, Mapuche, Yaghan, and Indians in
Chile (item 78).
It should be remarked that there is often even more uncertainty in these results as to the exact ethnic status of the
persons tested than in the results for North America and
Alexico.
These conclusions will be perhaps surprising to some. It
must be admitted that they are tentative only, but, I think,
justified on the basis of the data now available. I n any case,
the general fact would seem to be uncontrovertable, that it
can no longer be maintained that pure-blooded American
Indians were all originally in group 0. The exact classifkation of the many tribes not yet tested, and of many of those
listed here, must await further study. It is hoped that the
results and analysis presented here will be a further stimulus
for such study, and, by tending to bring our ideas of Indian
blood groups more in line with what is known about the
Old World, may help remove the prejudice which has been
associated in the minds of some anthropologists with the
idea of blood group studies.
It is tempting to comment on the striking differences between
the North and South American results, but this would be
risky until further studies are made. On the basis of what
results we have, it seems that B is absent in North America,
but present, though irregularly, in South -4merica. Whether
this will afford any clue to assist in reconstructing the history
of early migrations to and in America, remains to be seen.
I n order to focus the argument on a single issue, I have
deliberately omitted reference to the results of serological
absorption tests on mummified tissue (Boyd and Boyd, '37;
Wyman and Boyd, '37; Matson, '36), which indicate conclusions on the whole rather similar to the above. For the
BLOOD GROUPS O F AMERICAN INDIANS
233
same reason, nothing has been said about the blood groups
of the Eskimos, where also amounts of A impossible to attribute to European mixture are found, or about the newer
blood factors M and N. I n the latter connection it may be
remarked that there seems to be more uniformity in their
distribution among different tribes, so far as we know at
present. It is to be hoped that more studies of the M, N
frequencies will be made, for such work will enable us to
estimate independently the amount of European mixture
which map reasonably be assumed to be present in various
tribes.
It may be remarked finally that there is no longer any doubt
that the Australian aborigines possessed before European
contact a high percentage of A, which is again a reversal of
the ideas of the first workers in this field, and may serve to
increase our confidence in the above conclusions.
Thanks are due to colleagues who kindly criticised this
paper while it was in the course of preparation, and to
Dr.J. C. B. Grant for permission to use data still unpublished.
SUMMARS
A mathematical analysis of the data available on the blood
groups of the American Indians does not support the old
hypothesis that they were all originally in group 0, but discloses instead that the majority of North American Indians
probably possessed more or less a,though some may have been
pure 0, and the South American tribes, though some seem to
be pure 0, possessed in some cases A, in some cases B, and
perhaps sometimes both.
LITERATURE CITED
ALLEN,F. W., AND J. KORBERCited in Allen and Sehaeffer ( '35).
1935 The distribution of the liumnii blood
ALLEN,F. W., AND W. SCKAEFFER
groups among the Navajo and Pueblo Iiidiana of the Soutliwest.
Unir. N. Me+.Bull., Biol. Ser., IV, 2.
ALLEN, F. W., AND H. n. IARSEN 1 9 3 i Heredity of agglutinogens hf and N
among Pueblo and Blackfeet Indians. J. Immunol., XXXII, 301.
234
WILLIAM C. BOYD
ARCEL N U ~ ~Cited
A in Steffan and Wellisch (’33).
BERNSWN,
F. 1925 Zusammenfallsende Betraehtungeii iiber die erblichen Blutstrukturen des Menschen. Z. ind. Abstam. -u. verebungsl., XXXVII,
237.
1930 Ueber die Erblichkeit der Blutgruppen. 2. ind. Abstam. U.
vererbungsl., LIV, 400.
BOYD,Waa. C., A N D L. G. BOYD1937 Blood grouping tests on 300 rnumdea.
J. Immunol., XXXII, 307.
1939 Blood-group determinations upon the bones of thirty
C ~ H L P.
~ B.
,
Aleutian mummies. Am. J. Phys. Antbrop., XXIV, 361.
COOA,A. F., rn 0.DEIBERT1933 A study of the occurrence of the blood
groups among the American Indians. J. Immunol., VIII, 487.
DOWNS,C. M., H. P. JONES AND K. KOERBEB1929 Incidence and properties
of isohemolpsins. J. Inf. Dis., XLIV, 412.
FISHER,
R.. A., AND F. YATES 1938 Statistical Tables for Eiological, Agricultural, and Medical Research. Edinburgh.
GATES, R. R. 1929 Blood groups of Canadian Indians and Eskimos. Am. J.
Phys. Anthrop., XII, 475.
1936 Recent progress in blood group investigations. Genetica,
XVIII, 47.
1938 The blood groups and other features of the Micmac Indians.
J. Roy. Anthrop. Inst., LXVIII, 283.
GATES, R. R., Ah?) G. E. DAEBY 1934 Blood groups and physiognomy of British
Columbia coastal Indians. J. Roy. Anthrop. Tnst., LXIV, 23
GOLDEN,Q. 1930 Distribution of blood groups in South American Indians.
Lancet, CCXIX, 2, 278.
GOODNEB, K. 1930 Incidence of blood groups among the Maya Indians of
Yucatan. J. Immunol., XVIII, 433.
G U T , J. C. B. 1930 Anthroponietry of the Chipewyan and Cree Indians of
the neighborhood of. Lake Athabaeca. Nat. Mus. Canada Bull. LXIV,
Anthrop. Ser. no. 14. Ottawa.
1933 Progress in a n authroponietric survey of the Canadian aborigines. Fifth Pacific Sci. Cong. B 3, 2.
1936 Anthroponietry of the Beaver, Sekani, and Carrier Indians.
Nat. Mus. Canada Bi111. 81, Anthrop. Ser. no. 18. Ottawa.
1938 Private communication.
HIRSZFELD,
L. 1938 Les Groupes Sanguins. Paris.
HOGBEN,
L. 1932-1933 A matrix notation f o r Mendelian populations. Proc.
Roy. SOC.Edinburgh, L I I I , 7.
KAHN, M. C. 1036 Blood grouping of 336 Upper Aucaner Bush Negroes and
70 Alkuyana Indians in Dutch Guiana. J. Imniunol., XXXI, 377.
LANDSTEIN=,
K., AND P. LEVINE 1929 on the racial distribution of some
agglutinable structures of human blood. J. Immunol., XPI, 123.
LATTES,L. 1932 The Individuality of the Blood. London.
L ~ N EP,.,0. A. MATSONA N D H. F. SCRRADER
3935 Distribution of blow1
groups and agglutinogen hi among Indian ‘Blackfeet’ and ‘Blooil’
tribes. Proc. Soc. Exp. Biol. 6. Med., XXXIIT, 295.
BLOOD GROUPS O F AMERICAN INDIANS
235
MATSON, G. A. 1936 A procedure for the serological determination of bloodrela.tionship of ancient and modern peoples with special reference
to the American Indians, 11. Blood-grouping of mummies. J.
Immunol., XXX, 459.
1938 Blood groups and ageusis in Indians of Montana and Alberta.
Am. J. Phys. Anthrop., XXIV, 81.
MATSON, G. A., P. LEVINE
AND H. F. ~CHRADER 1936 Distribution of the subg r o u p of A and M and N agglutinogens among t h e Blackfeet Indians.
Proc. 8oc. Exp. Biol. & Med., XXXV, 46.
MATSON, G. A., AND H. F. SCHBADFZ
1933 Blood grouping among the 'Blackfeet' and 'Blood' tribes of American Indians. J. Immunol., XXV,
155.
MAZZA, S., AND I. FEANKE1927 Qrupos sanguheos de indios y de aut6ctonos
del norte Argentina. Yrensa MBd. Argent., XIV, 408.
MAZZA,SCH~~RYANN,
AND GUTDEIJTSCHCited in Steffan and Wellisch ('36).
MEZA Cited in Steffan ('32).
M o s s , W. L.,AND J. A. KENNEDY1929 Blood groups i n Peru, Santo Domingo,
Yucatan, and among the Mexicans a t the Blue Ridge prison farm in
Texas. J. Immunol., XVI, 159.
NIOO,C. 1926 A study of the blood groups among the American Indians. J.
Immunol., XI, 319.
PEARSON,
K. 1930 Tables for Statisticians and Biometricians. 3rd ed. London.
RAIIY, G. 1931 Lo8 grupos sanguheos de 10s Araucanos (Mapuches) J de 10s
Fueguinoe. Invest. y Prog., V, 160.
Cited in Lattes ( '32).
RIBERO Cited in Steffan and Wellisch '36).
RIDE, L., AND T. FUEUHATA
Cited in Ride ('35).
RIDE, L. 1935 Anthropological studies amongst North American Indians of
British Columbia. Caduceus, XIV, 205.
RIFE,D. W. 1932 Blood groups of Indiana in certain Maya areas of Central
America. J. Immunol., XXII, 207.
SNYDFX, L. H. 1926 Human blood groups: their inheritance and racial significance. Am. J. Phys. Anthrop., IX, 233.
1929 Blood Grouping in Relation to Clinical and Legal Medicine.
Baltimore.
SWFFAN, P., AND 8. WELLISCH 1933 Die geographische Verteilung der Rlutgruppen. Z. Raseenphysiol., VI, 28.
1936 Die geographische Verteilung der Blutgruppeii. Z. Rassenphysiol., V I I I , 38.
STEFFAN, P. 1932 Handbueh der Blutgruppenkunde. Munich.
STRANDSKOV,
H. H. 1931 A statistical study of the relative goodness of fit
of the two proposed theories of human blood group inheritance. J.
Immunol., XXI, 261.
WYMAN,L. C., AND WY. C. BOYD 1935 Human blood groups and antliropology.
Am. Anthropologist, XXXVII, 181.
1937 Blood group determinations of prehistoric American Indians.
Am. Anthropologist, XXXIX, 583.
AMERICAN J O U R N A L OF P H Y S l C A L
ANTHROPOLOGY, VOL.
XSV,
NO.
2
Документ
Категория
Без категории
Просмотров
0
Размер файла
886 Кб
Теги
group, indian, american, blood
1/--страниц
Пожаловаться на содержимое документа