Inflorescence structure and row number.
Two abnormalities have previously been described which affect row number in maize, each in its own particular way. (1) Multiplication, recently described by Cutler, produces two spikelets where normally there would be one. In its lowest expression it is responsible for the occasional kernel squeezed in between the regular rows of northern eight‑ and ten‑rowed flints. In its most extreme development it produces the crowded and apparently rowless ears commonly seen in parts of Central and South America. (2) Condensation (Anderson, Ann. Mo. Bot. Gard.) is a telescoping of successive internodes and is most easily analyzed in the tassel. In its extreme form it produces an eliptical or flattened, more or less fasciated ear. In its less extreme expressions it is responsible for most row numbers of 16 or above.
While these phenomena are not unknown in other grasses, as has been demonstrated by Cutler, they are both of them of a more or less teratological nature and it seemed probable that a study of the inflorescence structure in varieties of maize which have neither condensation nor multiplication might be illuminating. A special effort has been made to study such strains and, as anticipated, the structure of their inflorescences (tassels and ears) is much simpler than in other kinds of corn. Particularly as it concerns the central spike of the tassel, it does not seem to have been previously described. It is not spiral but whorled. There are two extreme types, those with whorls of two and those with whorls of three.
Old‑fashioned eight‑rowed flint corns exe an example of one extreme. Their central spikes are in whorls of two pairs of spikelets, each whorl bearing its spikelets at right angles to the whorls immediately above or immediately below. The uppermost tassel branches are also clearly in whorls of two. The other extreme type is found in certain persistently 12‑ and 14‑rowed strains of corn from South America and the Southwest. They have a structure similar to the eight‑rowed flints but the central spike has whorls of three pairs of spikelets and the upper portion of the tassel has whorls of three branches. In the Great Plains there are varieties with from 10 to 14 rows. When they are without condensation they show various mixtures of two‑whorled and three‑whorled.
The apparent spiraling of the central spike is due to the regular alternation of two patterns of spikelet position from node to node. In the eight-rowed flints, for instance, if the spikelets are on the north and south sides at one node they are on the east and west at the next, then the north and south again, and so on. In the 12‑rowed corns there is a similar alternation from positions A, C, E, to positions B, D, F, and then back again to A, C, E, producing a six‑ranked spike. Since each spikelet pair on the ear produces two kernels of corn the ear equivalent of a four‑ranked spike will be an eight‑rowed ear; for a six-ranked spike it will be a 12‑rowed ear. The structure of the tassel in these eight- and 12‑rowed races is almost transparently simple. The addition of a little condensation or multiplication, however, produces an organ so difficult to analyze that until these less complicated types had been studied the basic whorling was pretty completely concealed.
These observations allow us to put forward a series of hypotheses as to the various processes affecting row number in North American corn. They have already been tested genetically in part; further experiments are under way. The hypotheses are as follows:
There are at least four quite different characters which affect row number in maize. Each operates a different lever so to speak. (1) Maize is fundamentally either in whorls of two branches or whorls of three, or in various mixtures between these two extremes. There are indications that the genetic differences between the two‑whorled and the three‑whorled are multiple factorial.
In North America this basically simple difference is complicated by the almost universal presence of (2) Condensation. Preliminary genetic results suggest that this may be a single recessive gene, with a number of modifying factors which usually hold down the expression of this fundamentally teratological condition. In Central and South America (3) Multiplication is also an important factor in differences in row number. Nothing is yet known about its inheritance but various states of the phenomenon are known from very slight to very extreme. Except in an occasional inbred it is of little consequence north of Mexico. In addition to the above processes, row number can also be affected by the development or lack of development of the second floret as in Country Gentleman sweet corn and in various strains from South America.
These hypotheses can all be tested by orthodox genetic methods as soon as there are available multiple marker stocks which exhibit extreme values for the above phenomenon, viz., condensation vs. noncondensation, three‑whorl vs. two‑whorl, multiplication vs. no multiplication.