5. Tests of knotted, perhaps involving an inversion - Bryan (N.L. 1938, p. 5) reported Kn in this relation: br 7.2 f 27.0 Kn 24.1 Bm2. My last year's report (N.L. 1940, p. 17) was: an 22.5 Kn 25.2 Bm2 and an 22.6 Kn 9.6 gs. These 1940 reports were condensed from five-point tests including also br and f. In the five-point records given here those reported last year are combined with those obtained last summer.

It is obvious from these records that Kn is between an and gs and relatively near the latter. The recombination percentages for regions to the right of an are about those usually observed, but those in regions between br and an are far from normal. These differences in the two regions to either side of an are seen more readily perhaps when the data are assembled as three-point tests:

The data of the br-f-an array might indicate that the order of genes is not that given here. But the only order suggested by these data, on the basis of the usual criteria of three-point tests, is br-an-f. Since the chromosome-1 tester stocks employed in these tests are the same ones used in other tests (sections 1 and 2 of this report), no such assumption is tenable. It seems more likely that we are here dealing with a heterozygous inversion involving much of the region from br to an. This assumption is supported by the marked reduction in observed percentage of recombination, particularly in the f-an region, and by the appearance of more double crossovers than of singles in either region.