3. Most of our time and space this season was devoted to determining on which chromosomes are located the multiple‑factor segments which distinguish maize and teosinte. Relatively isogenic stocks, homozygous for one or more multiple-factor segments, were produced by crossing four varieties of teosinte with an inbred strain, backcrossing three times to the same inbred, and selfing. These were then crossed to a nine‑gene linkage tester and backcrossed to a second nine‑gene tester. The ears in these populations were then classified with respect to presence or absence of the multiple‑factor segments from teosinte. Such classifications are far from completely accurate, because the effect of the segments vary with the influence of several genes in the tester stock, especially j and g. Linkages can be detected, however, even when the classification is purely arbitrary, although exact crossing‑over percentages cannot be determined from these particular studies. The results of these tests are shown in the accompanying table. Analysis of the data was greatly simplified by the use of McBee punched cards which can be sorted with a simple, inexpensive tumbler.

Table I. Summary of linkage relations of the multiple‑factor
segments derived from four varieties of teosinte

Variety of teosinte	Number of segments	Linkage with chromosome number									Total number chromosomes tested
		1	2	3	4	6	7	8	9	10
Florida	1	‑	‑	+	‑	-	-	-	-	-	1134
"	1	‑	‑	+	‑	‑	-	-	-	-	1530
"	1	-	-	-	+	‑	-	-	-	-	1575
"	1	-	-	-	+	‑	-	-	-	-	1512
"	2	-	-	+	-	-	-	-	+	-	1512
"	2	-	-	+	+	‑	-	-	-	-	828
"	2	-	-	I	+	‑	-	-	-	-	1386
"	2	+	-	-	+	-	-	-	-	-	675
Summary	12	+	-	+	+	-	-	-	+	-	10152
Durango	1+	-	-	I	+	‑	-	-	-	-	567
"	1+	I	-	-	+	‑	-	-	I	‑	756
"	2	+	I	+	-	-	‑	-	-	-	1305
"	3	‑	‑	‑	+	-	-	-	-	-	1494
Summary	7	+	‑	+	+	-	-	-	+	-	4122
New	1	-	-	-	-	-	I	-	I	-	1539
"	1+	I	-	-	+	-	-	-	-	-	855
"	2	I	-	-	+	-	I	-	-	-	1575
"	2	-	-	-	+	-	-	-	I	-	1440
Summary	6	I	-	-	+	-	I	-	I	-	5409
Nobogame	1	-	-	-	+	-	-	-	-	-	1359
"	1	‑	-	-	+	-	-	-	-	-	765
"	2	-	-	-	+	-	-	-	I	-	1521
"	2	-	-	+	+	-	-	-	-	-	1602
Summary	6	-	-	+	+	-	-	-	I	-	5247
Grand Summary	31	+	-	+	+	-	I	-	+	-	24930

+ = Linkage

I = Indication of linkage

‑ = Independent inheritance

The important fact gained from this study is that the multiple-factor segments which distinguish maize and teosinte are located on chromosomes 1, 3, 4, and 9 in Florida and Durango teosintes. In Nobogame teosinte which had previously been shown to carry only three major segments, chromosomes 3, 4, and 9 are involved. In "New" teosinte chromosomes 3, 4, 9, and possibly 7 are involved. The remaining chromosomes appear to carry none of the major multiple‑factor segments which distinguish maize and teosinte. They are probably not lacking in genes which effect the various characters which distinguish the two species but these are either modifiers or segments too small to be detected by the methods followed in this experiment which depend wholly upon dominant or partially dominant effects.

It should be noted that chromosome 6 was not represented in the nine‑gene linkage tester. Previous studies on crosses of Florida teosinte with a stock including bm₁ on this chromosome gave no indication that it is involved in the four major segments.

The exact location of these segments and their length is yet to be determined. The segment on chromosome 1 shows very weak linkage with bm₂ and since previous experiments with Florida teosinte had shown one of the segments to be strongly linked with P at the opposite end it is probable that this segment involves part of the short arm of chromosome 1. There is some crossing over within the segment.

The segment on chromoaome 3 shows 25‑30 per cent of crossing over with A. This segnment is usually transmitted intact. Crossing over, if it occurs at all, is not readily detectable.

The segment on chromosome 4 includes the Su locus. There is considerable crossing over (about 30 per cent) within the segment.

Nothing is known about the position of the segment on chromosome 9, or the amount of crossing over which occurs within it.

The effects of the different segments are alike but not identical. All reduce the size of the seeds, and the diameter of the ear. All of them increase the prominence of the glumes and the number of ears produced on a single plant. At least two of these segments contribute very noticeably toward the reduction of number of rows of grain. In another experiment single segments were first rendered heterozygous by crossing with the original inbred strain, and the hybrid was then crossed with a second inbred to produce a vigorous and uniform F₁ in which approximately half of the plants were heterozygous for the segment. Ears from plants heterozygous for the segments average two rows of grain less than those which lacked the segments.

The segments have no discernible effect upon the pairing of spikelets or response to length of day. It is probable that they carry genes affecting these characteristics but that threshold limitations prevent single spikes from appearing at these levels.

The corresponding segments derived from different varieties of teosinte are similar in the nature and magnitude of their effects. In each case the segment on chromosome 4 is the most "potent." In each case this segment exhibits crossing over within the segment. Furthermore, a stock derived from Florida teosinte and homozygous for the segment on chromosome 4 is almost identical with a corresponding stock derived from Nobogame teosinte. Differences in teosinte varieties are attributable to: (1) Differences in the number of major segments; (2) the genetic nature of the maize varieties into which they have become incorporated; and (3) the probable presence of additional smaller segments or modifying factors.

We have some evidence that a single segment in heterozygous condition can increase yields appreciably, the extent to which this happens depending in part at least upon the kind of germ plasm with which it is combined. Hybrids involving some inbred strains are noticeably improved where small amounts of teosinte germ plasm are included.

It has so far been impossible to detect these segments cytologically. Stocks heterozygous for the segment on chromosome 4 occasionally exhibit a region of weak pairing on chromosome 4, but since similar regions are found on other chromosomes little significance can be attached to this. Apparently the segments are at least partly homologous to the corresponding regions of maize chromosomes so that there is no regular and distinct failure of pairing.

The new data seem to establish beyond any reasonable doubt the hybrid nature of teosinte. At least the varieties so far studied are nothing more than maize which has been contaminated by another species. The contamination is not a random one but involves multiple‑factor segments of four, or in the case of Nobogame teosinte, three chromosomes. These foreign genes must have come either from Tripsacumm or from a "pure" variety of teosinte now extinct or yet to be discovered.

P. C. Mangelsdorf

(Ed. note: In correspondence Dr. Mangelsdorf has written, "I have an abundance of seeds of several nine‑gene multiple testers and shall be glad to share it with anyone who wants some.")