This is an old and well-known variety. The seed is very small, and the plant in consequence is at first feeble.

The sugar per cent is fair,

the glucose rather high, and the non-sugar low.

ALAPORE JOWAR.

This is a new variety the seed of which was received from India this year. The canes are short and small, with heavy compact seed heads and dark green leaves.

In 1888, 150 plots were planted, each with seeds from a single crossed cane. In 1889, 250 similar plots were planted. In 1890, 355 plots were thus planted, and in all 755 plots were planted with cross-fertilized seeds, and many hundred analyses of the juices of hybrid canes were made to determine the possibility of producing improved varieties of sorghum by hybridization.

It appears that all the varieties can be cross-fertilized, and that a multitude of new varieties can thus be produced.

Not only do varieties which resemble each other and which flower and mature at the same time cross-fertilize, as for instance Kansas Orange and Early Orange, but very dissimilar varieties also cross-fertilize, as for instance, the small and early maturing Early Amber with the larger and later Link's Hybrid.

The principal part of the work has been done on crosses of the following varieties:

Amber and Orange.

Amber and Link's Hybrid.
Amber and White India.
Orange and White India.

Orange and White African.

Orange and Link's Hybrid.

Link's Hybrid and White India.
Link's Hybrid and Red Liberian.

Link's Hybrid and White African.

The first effect of a cross between varieties appears to be simply an alteration in the germs of the seeds which have been fertilized by pollen from another variety. This does not appear to influence or modify in any way the canes which produce the altered seeds. It follows from this that sorghum canes, grown from good seed, planted in close proximity to non-saccharine or inferior varieties produce as good canes for

sugar-making as when planted away from other varieties. But the seeds produced by the canes near inferior canes may be modified in the germs by pollen from the inferior canes, so that the canes produced in the next generation by these seeds may be variable in type, uneven in time of maturing, and inferior in quality.

The varieties of sorghum do not cross so readily as is popularly believed. In many cases at this station thousands of seeds of two varieties have been intentionally mixed and planted together in the hope of obtaining a desired cross. In some cases no crossed plants were produced; in other cases but a few crossed plants were found in the second generation, at which time the effects of cross-fertilization the previous year may be clearly seen. But while varieties of sorghum do not freely intermingle, yet cross-fertilization is effected under favorable conditions, and continuous planting of seeds which may contain but a few crossed seeds results finally in variable and mongrel seeds which produce canes which are uneven in ripening, uneven in sugar content, and not uniform in type.

The first effect observed, when cross-fertilized seeds are planted, is, in many cases, remarkably increased vigor of the crossed plants. The canes produced the first season after the cross-fertilization are often much larger and taller than caues of either of the parent varieties. The seeds and the seed heads are often remarkably large.

In looking over a field of cane it is safe to conclude that those canes which tower above the majority of the canes are crossed canes, or are stray canes from some larger variety. In selecting seed heads from a heap of cane tops, those which are remarkably large and fine it is safe to con. clude are from crossed canes or from stray canes of larger varieties.

It follows from this that the common practice of selecting the largest and finest seed heads for planting leads to wrong results in two ways: It is in many cases a selection of seed from crossed and variable canes, which is especially to be avoided, and it is in any case a selection of seed from canes which tend more to grain production than to sugar production, for, as a rule, the largest grain-producing canes have produced less sugar.

The extraordinary vigor, the unusually large growth of cane, the large seeds, and the heavy seed tops are usually found only in canes produced by seeds which were cross-fertilized the season before, that is, in the first generation of caues. In many cases crossing appears to act as a temporary stimulant, the canes in subsequent generations being no more than ordinary size. But there have been cases at this station where seeds cross-fertilized in 1887 produced canes in 1889 and again in 1890 which were larger than the canes of either of the parent varieties.

When cross-fertilized seeds are planted the canes of the first generation usually show remarkable variations in type, many showing an evident approximation to one or other of the parent varieties. In most

cases the departure from the typical character of the parent variety covers the widest range from the least to the greatest variation. The "seedling" canes often differ from any known varieties. When it is remembered that a single seed head usually contains from 500 to 1,000 seeds, and that these seeds may have been cross-fertilized in different degrees, the cause of these variations is partly explained. But it also appears that cross-fertilization also "breaks up" a uniform variety and induces variations in type, and sometimes causes departures from any known types, so that the crossed plants differ in foliage, in habit of growth, in color of seed top, in size, in earliness of maturing, in fiber, and in quality of juice. The variations in form are apparent to the eye; the variations in the quality of the juice are shown by analysis. In the canes produced by one crossed seed head, one may have 8 per cent of sugar, another may have 16 per cent, so that it is often impossible to make comparable analyses by taking samples of canes intended to represent the average value of the plot.

In some cases apparent reversions to some old form or variety are seen in canes produced from one crossed seed head. For instance, it is generally supposed that the Early Amber variety originated in this country. Mr. Leonard Wray, however, insists that he imported it to this country in 1851 under the name Boomvwana, which signifies red. The Amber is a black-seeded variety. In a dozen lots of Amber crosses reversions to a scarlet or bright red variety were seen, a color not derived from either of the parent varieties of these crosses, but probably a reversion to the original color of the Boomvwana variety.

With sorghum, as with other plants, crossing varieties causes variations in size, in type, and in qualities. Unusual qualities are sometimes found in crossed plants, and there are two methods known of producing improved varieties of plants from single crossed plants which have the desired qualities.

The most common method consists in multiplying the selected plant by bud propagation, that is, by bulb or tuber, by suckers, by graft, by runners, layers, or by cuttings. When a plant is thus propagated by bud, the new plant is simply a continuation of the plant from which the bud was derived; no new plant is formed and the plant is multiplied until it becomes a variety, without variations. It is now known that sugar cane occasionally produces seed, and these seeds crossfertilized have probably produced canes of unusual types, and these canes propagated by bud have perhaps formed the existing varieties of sugar cane. This method of propagation known and used by horticulturists allows at once the production of a uniform and improved variety from a single remarkable plant.

But this method appears not to be available for the production of improved varieties of sorghum, and the second method, propagation by seed, appears to be the only one practicable.

In this method cross-fertilized seeds are planted and from the vary