Abstract: Of 25 combinations between octoploid Trititrigia and tetraploid wheats, two F1 hybrids were obtained, both involving an 8X Trititrigia "Xiao-yian 7430", one with Triticum turgidum var. durum, another with T. timopheevi. The meiotic observations on the hybrids were made. Not only were there a lot of univalents, range from 9 to 19 in the hybrids with durum, 12 to 26 in hybrids with timopheevi, but there frequently existed remarkably complex configurations of multivalents as well. In 230 PMC’s of the latter hybrid observed, nearly all except 2 have one to five multivalents; the chromosomes associated in one configuration reach to six (sextivalent) and in a cell even a septivalent was observed, which must not be of homoeologous, but inhomoeologous nature. The mean configurations per PMC in both the hybrids are as follows: 13.5Ⅰ+11.1Ⅱ(3.04Ⅱ+8.05Ⅱ)+1.67Ⅲ+0.24Ⅳ+0.02Ⅴ+0.04Ⅵ in the hybrid with durum, and 19.8Ⅰ+7.01Ⅱ(3.50Ⅱ+3.51Ⅱ)+2.16Ⅲ+0.27Ⅳ+0.10Ⅴ+0.01Ⅵ+0.004Ⅷ in the hybrid with timopheevi. 6.3 univalents per PMC in the latter hybrid more than the former are clearly less than those of the hybrids between timopheevi and both durum and T. aestivum (even if the D genome’s chromosomes in the latter case were all regarded as univalents). Morever, a few multivalents (reach to pentavalent) occured occasionally, and especially in a tetrahaploid mutant happened spontaneously of the octoe ploid, bivalents (3.7 on the average) and even multivalents (as high as quadrovalent) were observed. All these data mentioned above indicate that there must have been duplications or repetitive fragments in the chromosomes of the octoploid.Observations on Anaphase I PMC’s of the hybrid with timopheevi showed that some and by no means all of univalents are orderly ranked on the equat orial plate of the spindles at the moment when chromosomes linked by chiasmata, say, bi-or multi-valents, begin to be segregeted. After the moment, the chromatids of the univalents on the plate start to divide and move to the opposite poles. Only being arranged near the the plate can the univalents be involved in this situation, in adition as another evidence the number of univalents (7.4 on the average) of Al cells on the plate is sharply less than that (19.8) of MI cells. That the behaviours of univalents near or far from the plate are so different is merely determined by their positions in the cell, as if there was a fansatic "polarity" across the plate as is proposed in a plausible hypothesis in this paper.To approach a way of producing secondary hexapoid Trititrigia, about which no report had been found before, a suggestion of reconstructing the genomes of 8X Trititrigia by 4X wheat has been but forward in the present paper. As an attempt the anthers of the hybrid between "Xiaoyian 7430" and T. timopheevi had been cultured in vitro, and several green tube-plants had been obtained.