Another dosage effect with regard to the length factors (L) of the ear was found in the progeny of the two lax - eared mutants. These mutants were tetrasomic, and the corresponding intermediate heterozygotes were trisomic. Crossings to the monosomic sortiment of "Chinese Spring" did not yield phenotypes similar to the hybrid progeny of normal "Koga II" x normal "Chinese" except for the disomic F₁ plants in the combination of the lax - eared mutant and mono - XXI. The latter types gave 21_II in metaphase I, while the trisomics in the same F₁ had 21_II + 1_I.

The last three mutants to be described here were produced by changes in the dosage of factors, influencing the strawlenght. Within a segregating progeny the normal phenotypes with 42 chromosomes had the longest culms, but the internodes remained shorter in trisomic intermediate forms and were extremely reduced in the tetrasomic mutants.

When summarized these results demonstrate gene dosage effects to be relatively frequent in hexaploid wheat. This is so much the more we favour the conception of micro - duplications in the chromosomes of all those mutants, that did not show aberrations in the microscop. With respect to the origin of the first squarehead mutant in the fields of the wellknown English farmer TAYLOR, it might be more reasonable to assume such a small duplication having no negative influence on the fertility than to trust in the polyfactorial hypotheses of squarehead inheritance postulated for so long time. Similarly our cytogenetic observations are thought to give the following clue to the evolutionary difference between Triticum aestivum subsp. vulgare and T. aestivum subsp. sphaerococcum. It is wellknown that SEARS was the first to locate the sphaerococcum factor of T. sphaerococcum on chromosome XVI by crossing different nullisomics of T. vulgare ("Chinese Spring") to T. sphaerococcum. Monosomic plants of the critical F₁ with 1 chromosome XVI of T. sphaerococcum were of vulgare phenotype, but the corresponding disomics in F₂ with 2 chromosomes XVI of T. sphaerococcum were of the sphaerococcum phenotype. Therefore SEARS argued the recessive sphaerococcum factor to be hemizygous - ineffective. Considering however, the same result in relation to the cytogenetical findings on our mutants, the following conclusions may be drawn: Just as sphaerococcoid mutants arose by indusing a duplication in the normal " Koga II ", T. aestivum subsp. vulgare may have a sphaerococcum factor on chromosome XVI, which phenotypically is ineffective in its normal diploid dose. In similar way T. sphaerococcum might be produced by a small duplication on chromosome XVI, i. e. in consequence of a dosage effect of the sphaerococcoid factor. Obviously this conception is confirmed by some crossing results of ELLERTON (1939) : In the F₂ of T. vulgare x T. sphaerococcum segregation was unifactorially in relation to the sphaerococcum characteristic. The heterozygous F₁ plants having 1 vulgare chromosome XVI (1 sp) + 1 sphaerococcum chromosome XVI (2 sp) were subsphaerococcoids comparable to our trisomic mutants.