1. In attempting to relate the chromosome circles which are observed in many species of Oenothera, to the extensive linkages which characterize this group of plants, Cleland makes two assumptions: (a) that homologous chromosomes are adjacent to each other in the circles, and (b) that chromosomes of maternal and paternal origin alternate with each other in a definitely fixed order. 2. The evidence cited by Cleland and others for the first of these theses is convincing, but not for the second. 3. In the absence of evidence regarding the arrangement of the chromosomes in the original diploid spireme following fertilization, it seems reasonable to assume that homologous chromosomes become associated in pairs which may lie in any plane with reference to adjacent pairs, and that when the spireme is formed, the maternal or paternal origin of the chromosomes has no influence in determining the order of arrangement of the members of each pair of homologues. 4. Even if it be assumed that the order of succession of maternal and paternal chromosomes in the original spireme is definite, they need not remain in the same order until the chromosome circles emerge in late meiotic prophase, since the looping of the chromosomes during "second contraction" provides for possible reversals in the positions of the homologous chromosomes. 5. A spireme in which each pair of homologues remains associated, but in which the order of maternal and paternal chromosomes is indeterminate, provides for the normal separation of homologues when adjacent chromosomes move to opposite poles of the meiotic spindle, and such separation is probably achieved even when adjacent chromosomes move to the same pole, if two other adjacent chromosomes in some other part of the circle move together to the other pole. 6. It also provides for a chance assortment at each pole, of chromosomes of maternal and paternal origin, just as is known to take place when the arrangement of the chromosomes is parasynaptic. 7. This being the case, there may be as many independent linkage groups as there are haploid chromosomes regardless of the grouping of the chromosomes into circles. The discovery of three linkage groups in Oe. Lamarckiana, in which the chromosomes are associated in a circle of twelve and one pair, is favorable to this view. 8. The telosynaptic arrangement of the chromosomes and their persistent cohesion into chains or circles, therefore, probably have no very fundamental genetical significance. 9. Irregularities in distribution of the chromosomes of the circles at anaphase are more frequent in such stable species as Oe. biennis, Oe. muricata and Oe. franciscana sulfurea than the number of visible genetical variations would lead one to expect if each such irregularity resulted in the exchange of genetically dissimilar chromosomes from one parental "complex" to the other. 10. Different species having like arrangement of the chromosome circles do not agree with one another in their genetical behavior, but seem to differ from each other rather on the basis of the number and nature of the lethal factors they possess. 11. Studies of Illick, Cleland and Hakansson have shown that at the time of second contraction in meiotic prophase, chromatin loops which probably represent pairs of homologous chromosomes show twisting and fusion which provide the possibility of transfer of chromatin particles or of whole sections of the chromosome from one homologue to the other, thus providing for crossing over within the linkage groups, notwithstanding the fact that the chromosomes are arranged telosynaptically. This sets aside Gates's (1922) contention that no mechanism exists for the production of crossing over in Oenothera. 12. The conception of the spireme here presented is just as hypothetical as is Cleland's view. Fortunately a decision between the two views seems capable of experimental attainment, especially through close cooperation between the experimental breeder and the cytologist.