508 
HEMATOLOGY 
(genotype Rrii) in matings to group i sheep of 
genotype RRii. 
That the R-0 system of sheep has some phys- 
iological significance is indicated by the re- 
striction of an alkaline phosphatase isozyme 
(zone B) to sheep of group 0.^2.23 But, in that 
connection, the more recent developments relat- 
ing the blood group locus M of sheep to high 
(HK) and low (LK) red cell potassium types 
are of more than usual interest because they es- 
tablish that antigenic determinants on cell 
membranes are involved in the transport of ions 
across those membranes. 
In 1966, Rasmusen and HalF* made the ob- 
servation that the genes v^^hich control the M 
blood groups of sheep^^ also control potassium 
levels. They observed that all HK sheep are 
homozygous for the alleles vi^hich code for the M 
determinants vi^hereas LK sheep are either het- 
erozygous or homozygous for the allele (s) 
vi^hich codes for the absence of the M determi- 
nants. 
The next obvious thing to do w^as to attempt 
to produce isolysins specific for the product (s) 
of the non-M allele (s). Both Rasmusen^^ in the 
United States and Ellory and Tucker^'^ in 
Britain v^^ere successful. At first those anti- 
bodies were designated anti-m but the name was 
wisely changed to anti-L, the reactions being 
limited, as they were, to the red cells of LK 
sheep. Now, in that same report, Ellory and 
Tucker recorded a most significant observation. 
They observed that when the red cells of LK 
sheep, either homozygous or heterozygous for 
the L determinant, are sensitized with anti-L 
there is a four-to-five-fold stimulation of potas- 
sium transport. In other words, they accom- 
plished in vitro the conversion of LK red cells to 
type HK by loading the L receptor sites with 
anti-L. This observation was soon confirmed by 
Lauf et al.28 as reviewed by Tucker.^" 
The significance played by antigenic determi- 
nants, or the molecules with which they are as- 
sociated, in regulating such things as the potas- 
sium level of red cells is yet to be appreciated. 
However, if the variation in antigenic determi- 
nants is largely associated with the immune re- 
sponse and thereby has evolved along with the 
evolution of that response, it seems likely that 
the effect of the 0 determinant of sheep on the 
appearance of an alkaline phosphatase isozyme 
or the L determinant on the level of red cell po- 
tassium may be purely coincidental and thereby 
of no unique selective value. 
SEROLOGY AND GENETICS OF ANIMAL 
BLOOD GROUPS 
In spite of the claim^^ that the great bulk of 
•blood group knowledge has been gathered from 
the results of simple (saline) agglutination 
tests, it may come as a surprise for some to 
learn that the foregoing firsts in animal blood 
typing were all brought to light by the use of 
hemolytic rather than agglutination tests. Ex- 
cept for the studies on pig blood groups and 
those of dogs (Table I) , the bulk of blood group 
knowledge, at least insofar as our larger species 
of laboratory animals may be concerned, has 
been gathered from the results of hemolytic 
tests. 
While simple in principle, none of the proce- 
dures used in blood typing is without its risks. 
Blood typing tests are extremely delicate, and 
require constant surveillance. Hemolytic tests, 
when properly performed, produce the most 
sharply defined and easily readable reactions. 
Moreover, the amount of antibody required to 
produce hemolysis is usually far less than that 
required to produce agglutination. For example, 
it has been estimated that two or three mole- 
cules of 19S (IgM) antibody is sufficient to pro- 
duce hemolysis of a sheep red cell,^^ but this 
may be a special case since it relates to the 
Forssman determinant. 
Complement is necessary for hemolysis and 
the key to the development of effective hemolytic 
tests is the use of an appropriate comple- 
ment. This is almost never fresh neutral serum 
of the species under study. For example, fresh 
neutral cattle serum is useless for cattle blood 
typing. It will produce stickiness and thereby 
agglutination of sensitized cattle red cells but 
such agglutination reactions are poorly defined 
with the average reagent used in cattle blood 
typing. Moreover, they are readily distinguisha- 
ble from regular agglutination because the 
clumps disaggregate as the complement be- 
comes inactive. 
