CLYDE J. STORMONT 
507 
pointed out the most effective method of devel- 
oping antibodies of diverse specificity for use in 
exploring blood groups, namely, isoimmuniza- 
tion. And it indicated the use of hemolytic tests 
where agglutination tests are clearly contrain- 
dicated, namely, those species like cattle, goats 
and sheep whose red cells are not prone to ag- 
glutinate even when saturated with multiple 
doses of blood typing antibodies. Unfortun- 
ately, neither of those observations became gen- 
erally appreciated until the 1940s. 
Another first which bears mentioning here 
was the discovery of what are nowadays re- 
ferred to as "inclusion" groups, such as the -in- 
clusion group BGK of cattle where blood factor 
K makes it appearance only in the presence of 
both B and G, although the latter two occur 
alone and together in the absence of K.^- Many 
inclusion groups of the order BGK have since 
been discovered, including cef of the Rh-Hr 
blood groups in man. They are of unusual inter- 
est in immunogenetics because they defy inter- 
pretations based on closely linked genes.^^ On 
the other hand, they connote cross or overlap- 
ping reactions among the isoantigenic determi- 
nants on red cells, and are readily interpretable 
in those terms. Blood factors B, G and K of cat- 
tle segregate as the elements of a series of re- 
lated determinants defined by multiple allelic 
genes S^gk^ ^bg^ and b, the latter cod- 
ing for the absence of B, G and K. Discovery of 
inclusion groups of the order BGK were of par- 
ticular importance for two reasons: (1) they 
brought to light the existence of determinants 
that are capable of binding antibodies of more 
than one specificity, and (2) they have helped 
considerably in the elucidation of such complex 
blood group systems as B of cattle.^* 
Although it is generally assumed that all 
antigenic determinants recognized on cell mem- 
branes are intrinsic properties of those mem- 
branes, there are some notable exceptions, in- 
cluding the Lewis determinants on human red 
cells. 1^ These exceptional determinants are sol- 
uble substances produced in the tissues, se- 
creted into the blood plasma and acquired by 
the red cells on contact with them in blood 
plasma. The first example of a red cell determi- 
nant of that order was J of cattle, as shown by 
both in vivo and in vitro experiments,^^ and 
this observation was followed by studies" 
which showed that the R and 0 properties of 
sheep red cells are also acquired from blood 
plasma. Aside from ascertaining the origin of 
certain antigenic determinants on red cells, 
these observations were of particular signifi- 
cance in accounting for the reasons why chi- 
meric twins in cattle, sheep and man may differ 
in their J types, R-0 types and Lewis types, re- 
spectively. It is clear from such studies that the 
tissues which give rise to these soluble sub- 
stances are not exchanged by the twins in utero. 
It is axiomatic in immunogenetics that any 
antigenic determinant, whether membrane 
bound or not, is inherited dominant in con- 
trast with its absence. There are a few excep- 
tions to that rule, the first bonafide example 
being the 0 determinant of sheep which is in- 
herited as a recessive to R.^^ 
It is also axiomatic that genes controlling an- 
tigenic determinants are expressed in any geno- 
type. Here, again, the first bonafide exceptions 
were provided by studies on animal blood 
groups. Utilizing anti-sheep 0 provided by the 
author, Rendel et al.^'^ observed that some 
sheep, rather rare in occurrence,^^ failed to 
manifest R and 0 substances on their red cells 
and in their plasma. They designated this third 
group in the R-0 system as group i and estab- 
lished that a gene, designated I, at a second 
locus, is necessary for the expression of groups 
R and 0. Thus, sheep of genotypes RRII, RRIi, 
Rrll and Rrii express R in their plasma and on 
their red cells, and sheep of genotypes rrll and 
rrIi express 0 in their plasma and on their red 
cells, whereas sheep of genotypes RRii, Rrii and 
rrii do not do so. This kind of gene interaction 
is referred to in genetics as epistasis, and the 
example involving groups R, 0 and i of the R-0 
system of sheep constituted the first example of 
epistasis involving antigenic determinants. 
Since then there have been others, for example, 
the A-0 groups of pigs^o and the ABO groups 
of man.21 The existence of epistasis is particu- 
larly important when interpreting the results of 
blood typing tests in connection with problems 
of questionable parentage. It is possible, as an 
example, for sheep of blood group 0 (genotype 
rrll) to produce nothing but group R offspring 
