26
In the cross (By, -3B4M4) (T-L-)v, X (1 +5B-M-)(T#Lt)V,*, vi" is more
frequent both in By- and in prototrophs. It is linked therefore to
(G,L). However, it is also linked to Lac, in the other linkage group.
There is, therefore, but a single linkage group, of which the only
suitable arrangement is: B.-(B,M)..Lac...V,..(T,L).., the order of

the factors within the parentheses being indeterminate from the present

datae

crossover theory

In an attempt to compute map distances from the available
data, a crossover theory must be used in which,unfortunately, no
correction for chiasma interference can be made. Such interference
conceivably may result in large discrepancies between true and esti-
mated values of map distance, particularly in the region (B,M) to (T,L).

As can be seen from the map, Table 6 » an interchange between
By and (B,M) results in B )4B4M+...; lack of interchange in B,-B+M+...
The ratio between these two types is 8:79. indicating a proportion
of interchange to total of 8/79+8 or 9.2%. With so small a distance,
correction for double or multiple crossovers would be negligible
compared to the experimental error. With the formula developed below:
tanh x = interchange: no interchange, a value of 10.1% would have been
obtained.

The estimation of the distance (B,M) - (T,L) requires a
detailed consideration of multiple crossing over. Absolute values
for the distances (B.M) - Lac; Lac - Vz 3 V,-(TL), or a, b, c,re-
spectively, are not available but only their relative proportions
as given by the ratios of the single-crossover types in Table 6 .

The values of a,b,c can however be estimated from their ratios,
Tag » Ty » Yo and the proportion of the "triple crossover class of
table ct, rq, since the trequency of multiple crossing over will de-

pend on the absolute map distances.
A recovered prototroph chromatid will fall into the classes
a..ed according to the distribution of crossover "breaks" in its
various segments. Since interchange between B,M and T,L is required
to produce a prototroph, only those chromatids with an odd number of
breaks in the region a+b+c will be recovered.

The map distance, x, may be defined as (100 X) the mean
number of crossover "breaks" in a segment. In the absence of inter-
ference, there should be a Poisson distribution of éhromatids with
varying numbers of breaks, the frequencies of 0,1,2,n, breaks being
given by successive terms of the expression? e* (1, x, x°/2!, xM/n!),
In this case, only chromatids representing the odd terms of this
expansion can be considered, their sum veing e"* (sinh x) which is
equivalent to 1 - @7&% . The sum of the even terms is e~* (cosh x)
whence the expression tanh x for the ratio of interchanges to non-
interchanges. It can be shown from the addition formula for tanh

KLtXo that this formulation is equivalent to Haldane's .

addition formula X,9 = *1 + Xo - 2K] Xo

The expression e~*sinh x also applies to the chances of
interchange in any segmental part of xX, Ge&.+ a; b, ce. We have then,
the four expressions following:

(e78sinhaa)(e7>sinn b)(e7°sinh c) = rg = .020
e-X sinh x

(e74sinh a)(e7°cosh b)(e7°cosh c) = Pe = ,264
e7* sinh x

and soa forth ~

x 2 -b =-¢ “ ~
Since e@ “.e ~,e ° = e* appears in both numerator and denominator of

each of these expressions, they can be cancelled out leaving only

the hyperbolic terms:

sinh a. sinh b. sinh c = .O020
sinh x

 
28

sinh a. cosh b. cosh c
sinh x

7
264 hula VEE

cosh a. sinh b. cosh c = .448
sinh x

cosh a. cosh b. sinh c
sinh x

268

While the solution of this system of equations would provide
a theoretically exact solution for a,b,c and x, we are here concerned
primarily with the estimation of x, and this can be more readily
obtained with the help or certain approaimations. In particular, a
may be taken aS Sgx , where Sg 18 Pa/ratlptre » i.6., the fraction
of tne" single" crossover types represented by a. This is not exact
insofar as the proportions of the types a,b,c will not be directly
related to the distances a, b, c in all cases, the contribution of
the triple-crossover, single-interchange, types being as the Srd
powers of the distances, etc., rather than the first power.

The result of this approximation is the equation in one
variable:

sinh s,x . sinh spx . sinh 8¢% = .020 , which can be solved

errs en ee ee ee

by successive approximations. The solution is x = .80, or 80 units
(morgans). a, b, ¢ are then <&lt, 36+, and 22 units respectively.
As a check on the approximation used, the result of sub-

etituting a = » = c = x/3 may be considered, This leads to the

equation:
3 -
sinh x/3 = ,.020, x = .75, which considering
“sinh x”
the crudity of the approximation is in good agreement.

On the other hand the application of the uncorrected formula
(x/3)9= -020 gives the result, still lower, x = .73. This cor-

x
responds to counting single-crossovers only in the a,b,c types, and

taking the three regions as equal in length.
29

The values already cited: 21+, 56+, and ey will be taken
as representing the best available estimate of the three regions.

Interference, by shifting the distribution of crossover types
towards the lower values, would be expected to diminish ra and there-
fore lead to a low estimate of x. However, interference would also
tend to cause a spacing of crossovers, increasing the likelihood that
where there are, for example, three crossovers, one shall be found in
each segment and lead to a "d" type. With a random distribution of
crossovers, there is only about 1 chance in 5 that 6 crossovers will
be distributed l:l:1; by interference this figure could conceivably
be increased to 1/1, which would compensate for a fivefold bias
against triple-crossovers compared to single-crossovers for a given
value of x. Interference has not yet been sufficiently analysed in
other organisms to permit of any more direct evaluation of the extent
to which these effects will cancel each other. It would clearly be
desirable to find other means of estimating these distances, perhaps
py the use of biochemical markers located in the left hand region of
the map. A comparison of the results might give direct information
concerning the possible role of interference.

Linearity

In constructing a map, and calculating distances, it has been
taken for granted that there is in E. coli a system of linear linkage,
such as nas been demonstrated quite conclusively in Drosophila, and
inferred in all higher organisms. What direct evidence may one bring
to bear on this question?

The method which one is forced to employ in hybridizing this
bacteriun introduces certain complications. The classical proot of
linearity is based on the additive character of distances, expressed
in morgans, between loci occurring within the same linkage group.

The determination of map distances is based upon a comparison between
50

parental andinew combinations of linked genes, as determined in the.
progeny of zygotes selected at random. In E. coli, on the other hand,
one is limited to the recovery of that recombination class in which
there has necessarily been an interchange between certain biochemical
loci, in the cases here discussed, between (B,M) and (T,L).

. The data analysed above, concerning the segregations of Lac
and V1 cannot be used for a demonstration of linear order without an
error of circular reasoning. This is shown by the indeterminacy of
interference which affects very vitally the linear additive properties
of adjoining crossover segments.

The bearing of the reversed crosses tabulated in Tables 3a 4
has already been mentioned. They illustrate the combinatorial charac-
ter of the segregation mechanism but do not specify it more closely.
For example, one might postulate that genes of bacteria are embedded
in an n-dimensional matrix, which is ordinarily conserved, but which
occasionally permits of a gene-for-gene interchange. This is equivalent
to the "Konversion" theory once proposed by Winkler (1952) as an
alternative to crossing-over theory in higher organisms, and which has
been revived most recently, in modified form, by Lindegren (1947). The
Konversion theory ismade untenable by evidence for the interaction of
different interchanges, for example, between bac and V as already
cited. The Konversion theory can be made to fit such results only by
making it experimentally indistinguishaple from the classical cross-
ever theory.

The interactions between interchanges also exclude similar
matricial theories where the wnits are perhaps not single genes, but
blocks ot them- for example a multilinear radial arrangement. At
least for the genes involved in such interactions, one is forced toa
conclude that they are in a continuous segment. Other genes, as yet

unstudied, or course might be shown to be placed on branches or
other bizarre modifications of the chromosome, but thus far no need
for such exceptions has arisen.

Additional, perhars more direct support for the linear
order of genes is provided by data on the segregation of Ve summerized
in Table 8. It will be noted that the segregations of Lac, V, and Ve
are congruous between the By- and By + classes (the latter in the sense
discussed on p. ++). In the totals of "prototrophs" isolated from
B-H-T+#L+By ¢Lac+V7"V65 Xx Beli¢T-L-B, ~LaceVy Ves one finds Lac- 78%,
Ver 82%, and v4® 36%, indicating that, as in previous experiments,
Lac ig linked to (B,M) and V, to (T,L). In addition, Vg is linked to
(B,M) somewhat more intensely perhaps than is Lac. Since linkage to
By is already eliminated, one would predict Prom these totals, on the
hypothesis of linearity, that Lac and V, should be linked, with
Ve to the left of Lac.

The data of Table 8 confirm this prediction. The parental
couplings of Lac and V, are -P and #s respectively. Of 137 Lac-, 154
were Vg"; of 39 Lac+, 29 were V,%. The order V, ; Lac, Vz is also
supported, since the four most frequent types are those corresponding
to single-crossovers on this basis, while they would include multi-
ple-crossover classes with any other order.

Since Lac and Va are sepregating, the totals for the four
combinations of these two factors can be compared with those of
_ previous experiments. A 4x2 table comparison with the corresponding
cross, Table 5, row 1, gives Ye? 19.6 for three degrees of freedom,
While this would pe an exceedingly poor agreement if a normal dis-
tribution obtained, an analysis of variance by means of the variance
ratio shows that p = .O05 that the discrepancy can be accounted
for in terms af the variance of the replicated experiments.

The first 3 factors tested, 5,M,T,L,B,,Lac, Vy, and Vg

have been shown to belong to the same linkage group. It is, there-
Se
fore, extremely likely that there is but one linkage group in E. coli,
the chances that another of the sames magnitude exists being 2-7, or
-CO8, There is no cytological evidence to suggest more than one
chromosome in E. coli. No other genetically investigated organism
has so few linkage-groups. Cytologically, the nearest analogue is
perhaps the compound chromosome of Ascaris megalocephalus v. univalens,
en = 2.

Attempts to Induce Aberrations

Using a chromosomal theory as a working hypothesis, it was
hoped that some verfication could be found by the study of types in
which the normal order of genes was disturbed. Since there is only
one chromosome (from the genetic evidence), the only types of re-
arrangements would be changes, leading to a series of inversion tyres.
It was thought that such types might be detected by genetical pro-
cedures, by virtue of their effect on crossing over. In particular,
the occurrence of an inversion in the region BL - (B,M) would be ex-
pected to have the effect of eliminating the recombination classes
involving interchanges in this region. In the cross B-M-T+L+EBy+ x
BeM4T-L-B,- this would be equivalent to the suppression of prototroph
recombinants; B,- types, however, would be recoverable, and allow the
investigation of the extent of the changes.

Preliminary attempts to find such aberration types have,
to date, been unsuccessful. The procedure was as follows:

Following treatment with nitrogen mustard or 20,000 r of
x-rays, cells of Y-40 and of Y-53 were incubated separately for 24
hours, to allow the separation of cells or nuclei that might have
been associated at the time of treatment. The cultures were then
streaked out on nutrient agar plates. Single colonies of Y-40
were picked and streaked across a nutrient agar plate. Streaks of

similarly treated Y-535 colonies were made from the opposite direction,
556
so that in the center of the plate, cells of the two types were
mixed, treated colony by treated colony. The occurrence of colonies
which would not interact to produce prototrophs, as detected by
plating into minimal medium, would be an indicator that the combina-
tion was heterogeneous for an aberration. Since in these experiments,
both "parents" were exposed to treatment, each plating was equivalent
to the testing of two chromosomes, for the occurrence of an aberration.
No marked variation in the yield of prototrophs was noted in tests
involving lel mustard and 28- K-ray-treated chromosomes. This can
scarcely be regarded as an adequate sample in view of the stringent
selection imposed by the technique, which might be expected to
eliminate any aberration types which are even slightly less vigorous
then the normal. This consideration is especially relevant in view
of the "hemizygous" condition of any aberrations in the probably
haploid vegetative cells. These studies will be continued.

How Many Segregants per Zygote?

In the experiments detailed in this paper, recombinants
were obtained from different cell types vhich were first exposed to
each other in an agar medium. Therefore, each prototroph recombinant
colony seen by the experimenter marks the site of formation of a
zygote. The question may immediately be raised whether there are
at that site other recomination classes which, by virtue of their
biochemical deficiencies, may not proliferate within the prototroph
colony on the minimal selective medium. This is equivalert to in-
quiring whether there is but a single viable product of meiosis
(as in megasporogenesis in many higher plants) or more than one,
as in ascomycetes. The solution to this problem would be of
special interest in relation to the possible occurrence of four-
strand crossing over. In addition, if an appreciable proportion

of prototroph coloniés consisted of two distinct segregation types,
34.
4¢ would be necessary to isolate these types for the collection
of segregation data.

There are at least three ways in which a zygote might yield
more than one haploia recombinant. Firstly, the zygote might be
capable of proliferation in the diplophase (or sporophyte), leading
to the occurrence of several diploid cells, each of which might under-
go meiosis independently, and by chance yield several segregation
types. Secondly, a single zygote might produce, after meiosis, in
addition to the prototroph, the complementary multiple mutant class.
Thirdly, in a system of four-strand crossing-over, there might be two
supplementary prototroph recombinants differing in the segregation
of factors such as Lac and Vy for which the diploid was heteray gous.

Obviously, the proper investigation of these possibilities
requires that one stringently avoid contamination of one colony with
another. For this reason, the cell-suspensions used were diluted so
as to yield only about 5-10 recombination colonies per plate.

Crosses were mace between Y-40 and Y-sd (B-M-T4+L4By,4Lac4Vy"
x Btl+?-L-B, -Lac~V, ©) on B,- containing minimal agar medium. As
already noted, about 90% of the colonies from such a cross are
BtH+T4+L+B,-. The theoretical complementary class would be B-M-'-L-B, +.
Because of its nutritional deficiencies, it could not be expected
to proliferate on the minimal medium even had it been produced after
meiosis. The possibility remains, however, that a few cells of
this constitution might still be present among the 10® or so B,-
cells of the predominant type ina colony. By plating such colonies
into medium lacking By but containing biotin, methionine, threonine
and leucine, the B,- cells would be suppressed, while the postulated
multiple mutant type coulda form colonies and be recovered.

The experiment just described was carried out, testing 5¢

colonies for their content of other cell types. In general, a
thininless colony could be shown to contain from 10-100 cells capable
ot forming colonies on the 6,l,T yL medium. However, in each case
investigated these have been shown to be indistinguishable from the
y-40 parental B-N-type, and must be presumed to arise from a sur-
prisingly low degree of contamination of the colony with these cells
fron the heavily seeded plate. A few colonies were found which could
be characterized as reversions from b,- to Bit These experiments
are, then, incjusive with respect to the occurrence of complimentary
geneotypes in the same colony. With appropriate stocks, not as yet
available, it should eventually be possible to menipulate the situation
so that the complementary type could be recovered selectively, ex-
cluding both parents and the dominant recombination class.

A search for supplementary types was conducted with the
game crosses, except that colonies appearing on By, ager were streaked
out directly on EMB-lactose agar to determine whether any of them
were heterogeneous for this factor. In some cases, a number of
4solated colonies from each FMB-test plate were then also tested for
homogeneity with respect to T1l- resistance. About 90 colonies were so
tested; only 1 colony was found containing both Lact and Lac- cells.
It is impossible to be certain that, with this low frequency, the
single colony which was picked actually was derived from two distinct
zygotes. These experiments cannot be considered as bearing critically
/ on the question of the occurrence of two- oP four-strand crossing
over because of the absence or information concerning the viability
of more than one meiotic product.
Diploidy.

The segregation of characters observed between prototroph
recombinants strongly suggests naploid condition of E. coli, with re-
duction immediately following the zygotic fusion. If this condition

could be modified so as to yield stable diploid variants, which,
36

using with haploid cells, might yield zygotes showing tri-somic

egregation, many questions concerning centromere relations, and the

umber of strands as melosis, as well as the dominance or recessive-

ess of particular characters, coulda be studied. Because of the

4fficulties of cytological examination, a genetic test was divised

hich, it was hoped, would detect stable diploid variants.
In 1913, Penfold, described a peculiar response of KE. coll

o sodium chloroacetate. Wild type strains appeared to be inhibited

yy this agent, but gave rise to resistant mutants, which appear as papill-

1e or button-like projections from the inhibited growth on mutrient

igar containing 0.1% sodium chloroacetate. The resistant mutants were

yeculiar. insofar as they lacked the capacity to form gas from glucose,

atthough abundant acid was formed.

These findings were contirmed with E. coli, K-12. In addition,

it was found that while the resistant mutant, Cla™, could rorm gas

from formate, it could not from pyruvate. The presently accepted

scheme for the formation of gas from carbohydrates by E. coli involves

the splitting of pyruvate to acetate, or similar Co fraction, and

formate. The formate is then decomposed to CO, +H, by the "formic .

hydrogenlyase” complex. Since the capacity to form gas from formate

is intact, while that from pyruvate is impaired, it may be assumed
that there is a correlation between Cla® and the enzymatic splitting

of pyruvate. While other interpretations are perhaps not ruled out,

this was adopted as the most likely explanation, and is the basis

for what follows.

on a priori ground, and from the work of Seadle and Coon-

radt on Neurospora heterocaryons. it is likely that the ability

to perform a reaction wiil generaliy be dominant to the inability.

Ola’ therefore should be dominant in a diploid heterozygote to Cla?

with respect to enzymatic function. If the correlation between
S7

enzymatic function and resistance is maintained, Cla® will also be
dominant with respect to resistance, 1.¢., the combination Cla*%/cla™
will still be sensitive to chloroacetate. Starting from the diploidised
wild type, Cla®/cia*, it will require two mutations to produce the
resistant type ClaT/cia™. On this basis, a diploidised E. coli should
yield mutations to the phenotype of resistance to chloroacetate only
with extreme infrequency compared to the normal haploid form. Since
ane need merely streak out cells of a suspected diploid type on
chloroacetate agar, and record the development of papillae, this
working hypothesis provides a possible tool for the detection of diploids.
Unfortunately, none have yet been found among some dozens of tests of
camphor- or acenaphthene~ treated material. This matter was discussed
primarily to illustrate a possibly very fruittul line for further re-
search, particularly from the point of view of the possibility of
cytogenetical correlations.
Transformation. Experiments designed to extract transforming factors
from cells of E. coli were mentioned in an earlier section. The fail-
ure or such experiments igs in line with the genetical properties of
the recombination system, linkage, etc., but cannot be regarded as
conclusive for the exclusion of aiffusible transforming factors. The
methods used may have been too delicate to extract appreciable
quantities or too rough to preserve what was extracted. There is,
however, a further type of genetic experiment which bears on the
possibility of transformation via soluble substances.

A glance at Tavle 7 shows that some "aultiply-transformed"
classes are more frequent than those involving changes of but one
or two loci of one of the parents, Un the transformation hypothesis,
this might be interpreted in terms of the non-uniform susceptibility
of different cells to transformation, so that wherever it takes

place at all, it is likely to affect several genes. Under these
o&8

conditions, one would anticipate that susceptible cells might be
influenced simultaneously by a mixture of transforming factors.

On the transformation hypothesis for the exchanges described in this
paper, it would be equivalent to predicting tri-parental recombina-
tions in mixtures of three genetic cell-types.

A sexual mechanism has rather different consequence. Among
other sexual organisms, biparental inheritance is the ruls, barring
the dubiously relevant exception of certain multi poric embryo-sac
types in the angiosperms. In mixtures of three genetic types, only
those types of zygotes may be inferred which result from pairwise
fusions of cells. Zygote formation is so infrequent that the coin-
cidence of successive fusions or the segregants of a Type 1 ¥ Typee
zygote with a Type 3 gamete has a negligible likelihood,

This critical point of difference was subjected to experi-~_
mental test in the following way. The same biochemical parents were
used as before, namely B-M-~-, and T-L-B,-. Lac and V, alleles were dis-
tributed among these parents in various combinations. For example,
the types BelM-Lac-VyF ; T-L-B,-bac-V,8; and T-L-B, -Lac+V,* were used.
Suspensions of these types were prepared and all three mixed together
in a manner analogous to that already described for normal, pairwise
crosses, From such combinations, prototrophs or B,-, otner factors
prototrophic, could be produced only by recombination between the
B-M-parent and one or the other of the T-L-B)- parents, By biparental
inheritance, in this case, only three otf the possible combinationsof
Lac ana V could appear among the prototrophs: Lac-V,"5 Lac-V,8;
Lac+V,™, If a ménage a trois were permissiole, however, the fourth
type Lac+V,5 should be found also.

By using different combinations or alleles, the experiment
may be varied so that a different class becomes the exceptional in

each case. In Table 9, the results of four experiments, so constructed
59
that a different class should be lacking on the basis of biparental
inheritance in each experiment, are set down. It will be seen that
no exceptional types appeared in a total of 628 tests. It may be con-
eluded that genetic factors from different cells are not freely
miscible, as would be demanded by the simplest versions of transfor-
mations. On the other nand, gene recombination is restricted in
any inetance to exchanges of genetic material between two cell types.

The results of this experiment are also a check on spontaneous
mutation as the source of what have been claimed to be recombinations.
On the spontaneous mutation hypothesis there should be no discrimina-
tion against the exceptional types which were not found in these 6x-
periments.

The genetics of bacterial transformations is still in an
exceedingly primitive state, and there is no information concerning the
occurrence of interactions in bona fide transforming systems, If
transformation is to account for the results of the present experiments
4t will have to fulfill the following conditions: a) non-independence
of factors, simulating a linkage group; b) potential capacity of
carrying all the genetic factors of the donor in a single parcel;

c) immiscibility of parcels derived from aifferent cells. From a
genetic point of view, such a transforming factor would be indistin-
guishable from a gamete, and its definition would be based on chemic al
properties only. It will be recalled, however, that Muller (1947)

has interpreted the pneumococcus transformation in similar terms:
"still viable bacterial chromosomes oP parts of chromosomes floating
free in the mediunm....these have penetrated the capsuleless bacteria
and, in part at teast, taken root there, perhaps after having under-~
zone a kind of crossing-over with the chromosome of the host.”

Further genetic work on transforming systems will be required

to substantiate this interpretation, particularly in view af the
40
relatively low molecular weight, 600,000, which has peen ascribed to
the pneumococcus factors. (Avery et al. 1944).

Other E. coli strains
In an attempt to find how generally the ability the recombine
49 distributed among bacteria, studies were made on two other strains

ot E. coli. These were designated B/r and L-15. B/r, obtained through

we

 

 

the courtesy of Dr. E. Witkin, is a radiation-resistant mutant of
strain B. Hoth » and B/r have been used extensively in studies on
mutation from phage-sensitivity to phage-resistance. L-15 is a strain
used by Roepke, Libby and Jones (1944) for the production of blochemical
wutations. A variety of biochemical mutants of L-15 were obtained through
the courtesy of Dr. Roepke. Double mutants of B/r, arzinine-methio-
nineless and histidine-p-sminobenzoreless.., were obtained from ultra-
violet treated material by previously aescribed techniques for the
tsolation of mutants, Lederberg and Tatum, (1946a) All combinations of
the T-L-B,- mutant, Y55, of K-12, and of the mutants of B/r and of 4-15
were made and plated into minimal medium as already described, In no
case was there any suggestion of the formation of prototropnhs within
L-15 or B/r mutants, between them, or with K-12. +t is recognized that
the conditions for recombination in those strains may differ from K-iz,
or that there mea be genetic conditions of mating type. ft was esti-
mated that recombination would have been detected had it occurred with
a frequency of not less than 107° of that found in K-12 mutants,
Serious cytological studies seeking to identify the zygote
in K-12 have not been attempted in view of the futility of attempts
to characterize and verify so rare an occurrence. The burden of this
investigation has been the verification of the recombination of genes
in a bacterium, and the elucidation of some of its genetic properties.
The way is open for considerable further work, using recombination as

a tool of genetic analysis, and to the more detailed picturization of
41

process in space and time.
Discussion

Although a transformational interpretation of these experiments
has not been excluded beyond any shadow of ceubt, it makes little
difference for most purposes whether one adheres to invisible zygotes
as against unextractable transforming factors. The techniques de-
scribed should be useful in either case toward the solution of genetic
problems in bacteria. Many of these are discussed in the Cold Spring
Harbour Symposium for Quantitative Biology, Volume 11, 1946, which
deals with the genetics of microorganisms. These problems include
the genetic nature of phenotypically complex mutations, the verification
Or reverse mutation as the bvasis of genotypic reversion, the site or
interaction of certain mutations involving glycolytic enzymes, the
genetic basis of antigenic variations, the verifivation of the "one-to-
one" theory of the relationship between genes and enzymes in bacteria,
and in general, any instance where it is required to test the allelism
or two or more genetic variations,

Genetic recombination has, of course, a far broader meaning in
biology than as a laboratary tool. The recombination of mutations is
a source of variation that may be of crucial importance in the evolu-
tion of new “adaptive peaks". While this statement, in its general
terms is indubitable (see Dobzhansky 1941) recombination can only
effect the reshuffling of preexistent mutations, Concerning the natural
historical ai gnificance of the latter for bacteria, we are in a state
o1 woeful ignorauce, so that we are hardly in a position to discuss
the significauce o: bacterial recombination in concrete terms. To this
must be added the caution that geritic combination was found, luckily,

alfboyah
in one E. coli strain, and not in two others which were studied.
42

It way be wondered at that the apparent recombination rate is
so low. It will be recalled that about 1u7© of the cells insSoculated
ati the cross B-M-T+L+B)+ X B+M+T-L-By- showed interchange in the
regiou (B,M) - (f,L). Since the estimated map distance is 8U
units this 1s also the correct order of wagnitude of the fusion
process. However, this is possibly not to be ascraped to any
sexual imperfections of the colon bacterium, but to the method of
enumeration. It seems likely that an anal gous comparison of the number
ot somatic and generative cells in a nigner plant, or the ratio
of perithecia to total nuclei in a fruiting culture of Neurospora
would not give very different ratios. It 1s also possible that the
optimal conditions ror zygote fommation and germination have
not yet been achieved ahd that by special procedures the rate of
fusion may be wcelerated to the level where there might be some
hope of trapping 1t 1n the field of the microscope.

Since the mutants used in the recombination experiments were de-
rived trom the same wild-type strain, there can be no question of
genotypic mating type determination. The failure of two E. coli
strains to exhibit recombination might conceivably be ascribed
to genetic heterithallism, such as has caused many fungi to be
classified as "imperfect". The chances of finding the appropriate

mates are of course very slim, but should not be entirely overlooked.