1. A pericentric inversion in chromosome 9.

In the F₁ progeny from X‑rayed pollen there was detected in a single plant a long inversion in chromosome 9, which involved approximately .7 in cytological length of the short arm and .9 of the long am. The heterozygote gave about 25% aborted pollen. Cytological determination of the inversion could be made out only from pachytene configurations which, in heterozygous form, assumed a typical large loop. One of the inversion breaks was found closely associated with the wx gene. As in the manner of translocations in maize, the inversion was followed in linkage test by the partial sterility of pollen which behaved in outcrosses like a dominant gene located at the break point.

Linkage relations between the inversion break in the short arm (hereafter designated In) and wx, and between the gene sh for shrunken endosperm and In, were tested first in separate crosses. The data presented in the following table were obtained in each case from the type of testcross in which the female parent was the double recessive.

Table 1.

Year	Family	Constitution of male parent	Non-crossovers				Crossovers					Crossing over %	Total plant
			++	+-	-+	--	++		+-	-+	++
1947	102‑103	n* Wx		78	87		3				2	2.94	170
		In wx
1948	303‑304	n wx	75			64		7		2		6.08	148
		In Wx
1948	307‑308	n wx	96			87		5		4		4.68	192
		In Wx
			171	78	87	151	3	12		6	2	4.57	510
1949	442‑451	Sh In	170			86		4		7		4.12	267
		sh n
1949	452‑457	Sh In	92			95		8		4		6.03	199
		sh n
			262			181		12		11		5.08	466

* The symbol, n, represents the normal arrangement of chromosome 9 and it is treated as a recessive gene.

Obviously, averages of 4.57 and 5.08 can be taken only as relative measures of the respective linkage value for In‑wx and sh‑In since reduction of crossing over in these regions, as caused by the heterozygous inversion, will be evident in a three‑point test.

The inversion break in the short arm of the chromosome was placed between the genes sh and wx. Evidence for this sequence was given by the fact that, as a result of any single crossing over within the inversion loop in the inversion heterozygote, gametes formed with the chromosome deficient for the distal .3 of the short arm (and at the same time duplicate for a small piece of the long arm, hereafter designated as df 9S, dp 9L) were tested to be void of the allele of the sh gene. Part of the counts made in 1947 from twelve crosses of the type,

	Sh In	x		sh n	,
	Sh n

gave 77 shrunken kernels out of a total of 3180 seeds on the ears of the immediate cross, the frequency being 2.42%. Twice this figure, that is, 4.84%, was taken as the total frequency of the deficient‑duplicate gametes survived since the complementary class, designated as df 9L, dp 9S, should occur and transmit with equal frequency through the female side. Accompanying with cytological demonstration of a df 9S, dp 9L chromosome paired with a normal 9, the plant grown from such shrunken seeds gave 50% pollen sterility and produced all shrunken seeds when it was further tested by the recessive stock. It seemed evident, therefore, that the sh gene was located in the non‑inverted region of the short arm.

The data of a three‑point test, in which the heterozygote

Sh In Wx

sh n wx

was used as the male parent, were available in 1949 from seven cultures; lumped numbers being given as follows:

	Non-crossovers		Crossovers at region
	0		1		2		1.2
Family	ShInWx	shnwx	Shnwx	shInWx	ShInwx	shnWx	ShnWx	shInwx	Total plants
411‑437	266	225	8	12	8	3	4	5	531

Crossing over percentage: sh‑In = 5.46 In‑wx = 3.76

Linkage values for In‑wx and sh‑In obtained in this test were fairly close to that observed in the separate tests. Taking 30 as the map distance of nomal chromosome between the loci sh and wx, the heterozygous inversion obviously caused a marked reduction of crossing over in the region concerned; the difference (30‑5.46+3.76=20.78) being twice as much as the observed value of the heterozygote. It remained to be seen whether the reduction was greater or less in the sh‑In section than in the In‑wx region.

Another test was made to try to place the inversion break more precisely on the chromosome map with respect to the locus of bp, the gene for brown pericarp, which was known to be located midway between sh and wx. The bp material obtained from the Maize Co�p bearing the pedigree number 43‑163 (2)� was supposed to be a c sh bp wx tester. It was, however, segregating for brown and colorless pericarp plants, apparently due to the segregation of P gene. When the inversion stock of the constitution p Sh In Bp was crossed on to the brown tester, all the F₁ plants possessed red pericarp, and the inversion heterozygotes were selected again for crossing to the tester in both ways. In the progeny of the test, only red and brown pericarp plants were included in the counts for linkages since in the absence of P the phenotypes of Bp and bp were indistinguishably colorless. In table 2 the data represent the results of three types of testcross, namely,

(1)		P	x		P sh n bp	,
		p sh n bp			p Sh In Bp
(2)			x		P sh n bp	and
		p sh n bp			p Sh In Bp
(3)		P sh n bp	x		P	.
		p Sh In Bp			p sh n bp

Contrary to expectation, segregation of the pericarp colors was found to be almost independent of the inversion break and of the sh gene. The general tendency of the segregation seemed quite unique among the families. Even with inereased size of population, the result probably would not be changed to the other extreme. It seemed very likely that either we have dealt with another gene such as A^b or a^p, in addition to the presence of bp, which determined the production of pericarp pigmentation to the similar effect, or the bp gene might be located quite a distance away from the break in the short arm. Nevertheless, neither of the possibilities has been confimed at present.

In this connection, it is necessary to mention a test of the deficiency‑bp relation. Plants grown from the shrunken seeds obtained from the cross,

m	p	Sh n Bp	x	P	,
		Sh In Bp		p sh n bp

were classified for their pericarp colors. The data are summarized in table 3.

Table 3.

		Parental constitution				F1 plants and pericarp colors
Year	Family	Plant	Female		Male	Red (P Bp)	Brown (P bp)	White (p-)	Total
1948	344	102		Sh n Bp	p sh n bp	0	0	2	2
			p	Sh In Bp
	347	102	"		"	0	0	6	6
1949	572	307	"		P	1	0	1	2
					p sh n bp
	574	307	"		"	1	0	3	4
	577	320	"		"	1	4	0	5
	578	320	"		"	5	0	3	8
	579	320	"		"	2	0	3	5
	580	320	"		"	0	1	0	1
	581	321	"		"	1	2	0	3
	582	321	"		"	0	0	3	3
Total									39

Table 2.
Frequencies of plants in families

		Type cross 1						Type cross 2					Type cross 3
Crossing- over region	Phenotypes	442- 443	444- 445	446- 447	448- 449	Total	%	452- 453	454- 455	456- 457	Total	%	460- 461	462- 463	Total	%
0	P Sh In Bp	20	16	11	22	69		5	7	7	19		39	46	85
	P sh n bp	10	8	2	11	31		8	5	4	17		19	16	35
1	P Sh n bp	1	0	0	0	1	4.76 (4.12)*	1	1	0	2	6.97 (6.03)*	0	1	1	2.81 (2.74)*
	P sh In Bp	0	2	0	2	4		1	0	0	1		3	0	3
2	P Sh In bp	16	14	10	13	53	44.45	5	7	12	24	54.65	36	37	73	50.20
	P sh n Bp	5	12	1	9	27		9	9	2	20		54	15	49
1,2	P Sh n Bp	1	0	0	0	1		0	1	1	2		2	0	2
	p sh In bp	0	1	0	2	3		0	1	0	1		0	1	1
		53	53	24	59	189		29	31	26	86		133	116	249
sh-bp							45.00					54.65				50.60
	p Sh In -	11	14	9	14	48		18	15	16	49		31	27	58
0 or 2
	p sh n -	5	16	1	6	28		33	22	3	58		10	9	19
	p Sh n -	1	0	1	0	2		3	0	1	4		2	0	2
1 or 1.2	p sh In -	0	0	0	0	0		2	0	0	2		0	0	0
		17	30	11	20	78		56	37	20	113		43	36	79
	Total	70	83	35	79	267		85	68	46	199		176	152	328
	Unclassified	4	11	10	8	33		26	12	13	51		20	23	43

* Percentage of crossing over between sh-In when p- plants were not included.

While most families gave expected results on the basis of the supposition that the gene Bp was located in the inverted region, the appearance of brown plants in the families 577, 580 and 581 seemed difficult to explain on the same assumption. However, it was noted that families 577 to 582, inclusive, came from plants of different origin. The inconsistent result in 1949 might again be due to the involvement of another gene or genes in the latter cultures.

The transmission of a deficient chromosome through both sexes was tested against normal ones. It was found that in progenies of six cultures, consisting of 266 plants, no single gamete with the df 9S, dp 9L or the df 9L, dp 9S chromosome could go through pollen. On the other hand, deficient‑duplicate gametes were successful in competition with the normal gametes in a ratio of 1:6 (actually 17:127) to transmit through the female side.

In a certain setup, i.e.,

– Wx		,
	sh n wx

crossing over between sh and wx under the condition of heterozygous deficiency could be used as a check on the value of the heterozygous inversion. The value was determined to be 9.3% which was fairly close to what was observed in a previous section.

Different combinations of the chromosome with normal arrangement: typical inversion; df 9S, dp 9L and df 9L, dp 9S were investigated by cytological study for the purpose of determining the role of the centromere in synapsis. General observation revealed that the most frequent type of configurations was those resulting from the initial pairing of centromeres. Occasionally, a type was also found in the same material in which the terminal regions started to pair first, resulting consequently in homologous synapsis of the co‑regions. Non‑homologous synapsis without centromere pairing was also observed but was least frequent.

Material with a homozygous inversion has been made available for linkage study in this condition. (Acknowledgments are due to Dr. L. F. Randolph and Dr. E. G. Anderson for their generous help when the work was undertaken at Cornell University and California Institute of Technology.)

C. H. Li