2. Genetics of the B chromosomes and their derivatives. The B chromosomes are by no means genetically impotent as was formerly believed and is still being reiterated in current literature on maize cytogenetics. It is true that in small numbers they appear to produce no discernible effects; they are transmitted more readily than any known A chromosome fragments through both pollen and egg and their presence in genetic stocks seems not to have interfered with genetic analysis of mendelizing characters. But this does not necessarily mean that they are genetically inert or devoid of hereditary potentialities. In summarizing my data on the behavior of the B chromosomes that have been accumulated over a period of years in attempts to solve the enigma of their origin and fundamental nature, there are some rather interesting conclusions that can be drawn with reasonable assurance that they may mean something.

Although individual plants with relatively few B chromosomes are indistinguishable from their no B sibs, higher numbers of B chromosomes produce marked effects: More than 13-15 cause some reduction in fertility; more than 23-25 cause a marked reduction in both fertility and vigor; more than 30 occur rarely and the plants are very weak, produce mostly aborted pollen and set little or no seed.

In reciprocal crosses of plants with 1 B x 0 B, the B chromosome is transmitted about equally well by the pollen and egg to about one-third of the progeny. Exceptional plants with 2 or more Bs appear in these crosses more frequently when the B is carried by the pollen parent.

Reciprocal crosses involving 2, 3, and 4 Bs with no B plants are markedly dissimilar: when the Bs are carried by the seed parent, the numbers in the progeny tend to be intermediate between the parental numbers, but when they are carried by the pollen, the 0 B, 2 B and 4 B classes are predominant. This was true of both meiotic and somatic counts, the total number of individuals involved in these crosses being 398.

The B chromosome plants do not breed true for any given number of B chromosomes, regardless of whether the number is odd or even. When selfed, or when plants with the same number of Bs are sib crossed, less than one-third of the progeny have the parental number of B chromosomes. Various numbers are represented in the populations, the mean number being approximately the same or slightly less than the parental number for plants with from 1 to 17 B chromosomes. The total number of plants studied in these selfed and sib-crossed progenies was 988.

Numbers higher than either parent appeared frequently in crosses between plants with different numbers of Bs ranging from 1 to 20, but in the progenies of plants with more than 20 Bs they appeared less frequently. The mean number of Bs in the progenies of plants with from 1 to 10 Bs when intercrossed was essentially the same as the mean parental number; with higher parental numbers whose means ranged from 11 to 20.5 the mean number in the progeny was less than the parental mean by from 10 to 30 percent. These data were from 65 cultures which included a total of 983 plants.

Irregular assortment in meiosis, somatic nondisjunction and double division in somatic mitosis possibly due to irregular timing of centromere division, are some of the characteristics of B chromosome behavior responsible for the extreme variation in number observed in the progenies of B chromosome plants. Although the number of Bs in an individual plant is not necessarily the product of the contributing gametic numbers since changes in number may occur in autogency due to mitotic irregularities there is little evidence of selective elimination of gametes except among very high B chromosome plants. There is no evidence from these experiments on the breeding behavior of the Bs to support the contention of Darlington, presented in a recent discussion of "the activity of the inert chromosomes" (sic) in maize, that there exists a population pressure maintaining an equilibrium distribution of the B chromosomes at relatively low levels in different stocks. In fact the results suggest that higher numbers than are present in most natural populations would readily be tolerated. It seems quite possible that the B chromosomes are on the increase in at least some varieties of maize.

No disturbed ratios were obtained from F and backcross data involving B chromosome stocks crossed with 43 known genes distributed throughout the 10 linkage groups. The linkage relations of these genes are indicated on the accompanying map in which the tested genes are underscored. This map also includes tentative assignments of centromere positions based on information kindly furnished by Anderson, Rhoades and Burnham, the more definitely placed centromeres being represented by an oval drawn with a solid line and those less definitely placed being similarly represented by a dotted line. Disturbed ratios have been obtained with the gene sb, together with some evidence that the reduction in the number of recessives in the segregating progenies was proportional to the frequency of the B chromosomes. (See also Shafer's discussion of sb ratios in this News Letter) This would be expected if the B chromosomes carried the normal Sb allele. Unfortunately the linkage relations of sb are unknown.

These gene tests involving the B chromosomes have an important bearing on the fundamental question of the origin of the B chromosomes. If the centromere positions indicated on the linkage maps are even approximately correct, it is apparent that the tested genes giving undisturbed ratios in the presence of B chromosomes are distributed among 17 of the 20 normal A chromosome arms. Only 3 arms, the short arm of 8 and 10 and the long arm of 9, do not include at least one tested gene. If a test of one or a few genes were sufficient to exclude a particular chromosome arm from further consideration as the source of the B chromosome, the problem of the origin of the Bs would be much simplified, but in my opinion such tests would not be sufficient. It is altogether possible, in my opinion, that only part of a particular arm is represented in the B chromosome. For example, it might consist of an A chromosome centromere plus some adjacent euchromatin, but not necessarily all of the euchromatin of any particular arm, and in addition heterochromatin from the same or some other chromosome. This suggestion as to the possible mode of origin of the typical B chromosome may seem unnecessarily involved. However, there is a rapidly accumulating body of evidence that the chromosome is not as stable a unit as it was once thought to be. In fact it is surprising that chromosomes maintain any individuality whatever as separate and distinct morphological entities for extended periods of time in the light of the numerous types of reorganization to which they are subject. Furthermore, the typical B chromosome has a distinctive prophase morphology unlike that of any one region of similar length among the A chromosomes ordinarily present in existing types of maize. This is not an off-hand statement based on casual observation, but is the conclusion arrived at after making a very critical survey of the meiotic prophase morphology in well over fifty varieties of maize representing all of the known types of flour, flint, dent, pop and sweet corn, a survey that was conducted primarily to throw light on the origin of the B chromosomes. This does not mean that there may not be in existence today types of maize containing an A chromosome or segment thereof that is identical with the B chromosome. Or it may be that such a chromosome existed in primitive strains of maize that are no longer in existence. The fact that the B chromosome ordinarily does not synapse with any of the A chromosomes suggests that it is not of recent origin, but synaptic behavior alone should not be considered as proof of this assumption.

There is the further possibility that hybridization with relatives of maize may have been involved in the origin of the Bs, but in my opinion the possibilities of a more direct mode of origin are by no means exhausted.

In a further search for clues to the origin of the Bs, it would seem highly desirable to examine additional types of maize especially from regions where primitive stocks may still be in existence. Also more extensive tests of known genes should be made in the search for alleles of B chromosome genes; possibly Sb is one such allele, but additional cytological and genetical tests are needed to establish this. If the suggestion made above concerning the origin of the Bs is valid, and if there is a tendency in maize as in Drosophila for heterochromatic regions to be populated with fewer genes than are the euchromatic regions, the best chance of finding alleles of known genes in the B chromosome would be to test especially genes lying near the centromeres in the linkage maps. These genes may actually be an appreciable distance cytologically from the centromeres. But if the proximal euchromatic region of the B is in approximately the same relative position with reference to the centromere that it was in the A chromosome from which it originated, some of these nearby genes should be represented by alleles in the euchromatin of the B, which constitutes approximately one-third of the total length of the chromosome. A certain number of these nearby genes have already been tested as indicated on the linkage map. An especially good test involved chromosome 5 in which Rhoades' data from his telocentric fragment has given us the best evidence we have of the location of a particular centromere relative to neighboring genes. His evidence tells us that the closely linked genes, bm and bt, are definitely on opposite sides of the centromere. These two genes, as well as a2 in the short arm and bm, pr and v2 in the long arm of this chromosome gave normal backcross and F2 ratios in the presence of B chromosomes. Thus these tests would seem to exclude the possibility that the regions in which they are located are involved in the makeup of the B chromosome.

A notable characteristic of the B chromosomes is that they are like the A chromosomes in being susceptible to breakage, with the resultant loss of acentric segments of chromatin or rearrangement of parts. But there is this distinction that the supernumerary B chromosomes can undergo a greater variety of such morphological changes than can the A chromosomes without deleterious effects, and their monocentric derivatives can be readily maintained in culture for further study. Over a period of years a considerable number of such B chromosome derivatives have arisen in my stocks, the first of these being the C chromosome that was described back in 1928. Since most of these elements have been detected in root tip figures being examined for chromosome count, they have been grouped for convenience in four reasonably distinct size classes or types, based on their appearance in the somatic metaphase. These include (a), the C type that is somewhat shorter then the B chromosome but definitely elongated in contrast to (b), the D type that is essentially spherical with a diameter roughly equivalent to the diameter of an ordinary chromosome, (c), the E type that is of approximately the same size as the undivided satellite of chromosome 6, and (d), the F type that is distinctly smaller than the E type and in fact is only slightly above the lower limit of visibility of the photomicroscope.

On the basis of this classification there can be no additional new types of still smaller B chromosome derivatives, at least not until the electron microscope is utilized in the study of chromosomes. (Incidentally, this series of chromosome types from B to F, if interpreted in the reverse order, makes a very convincing demonstration of the de novo origin of chromosomes.) In the meiotic prophase morphological distinctions within those size groups can be detected and may be classified accordingly.

The B chromosome derivatives are proving very useful in studies of the relative genetic potency of different parts of the B chromosome. Data are available at the present time which suggest that the sterility-inducing effects of the B chromosome are to be attributed to factors localized chiefly in the proximal euchromatic region of the chromosome.

There is some evidence that other mutant derivatives of the typical B chromosome, such as extensions of the long arm or additions to the rudimentary short arm, occur from time to time, but these are less easily detected in somatic figures because of their greater similarity to the shorter A chromosomes.

The occurrence of distinctly dibranchial B type chromosomes in maize has been described from somatic figures by Darlington and others in recent years. But in these cases the position of the centromere has very probably been misinterpreted. The typical B chromosome when viewed in somatic metaphases often exhibits what appears to be a subterminal constriction, especially after fixation with fluids that shrink the chromosomes. This is not a true centric constriction but is actually the weakly chromatic region between the proximal knob adjoining the centromere and the distal heterochromatic portion of the chromosome. This interpretation is quite obvious if one is familiar with the pachytene structure of the B chromosome and follows the transformation accompanying the shortening of the B chromosome during the late prophase and early metaphase of the first microspore division where the distinction between euchromatin and heterochromatin in these stages is clearly apparent in good preparations. Many pachytene figures of the typical B chromosome do, however, show the presence of a rudimentary short arm consisting of a very few small chromosomes. This arm is often folded back against the proximal knob on the opposite side of the centromere, thus making the centromere appear truly terminal.

L. F. Randolph