satisfactorily still remains to be determined.

As to Natal grass, I have already mentioned that this succeeds better on the poorer and drier pine lands than any other grass yet introduced. Thus far it has been exploited purely as a grass for market hay. On this basis many hundred acres were planted in Lake County and elsewhere. Grass culture purely for market hay is a very precarious proposition. The proper agricultural economy is grass for live stock, selling only the surplus to the market. Notwithstanding the very large acreage planted to Natal, I have been quite unable to secure satisfactory data as to its value for pasturage, measured in carrying capacity and satisfactory gains. It seems to me, from the slender data I have been able to secure, fairly probable that Natal will prove a valuable grass for combined hay and pasture on the soils to which it is so well adapted, but of course it can hardly be expected to yield enough to justify the extravagant prices paid for land planted to Natal.

Prairie Lands.

On the prairies of Florida there is much better natural pasturage than in the piney woods, and, indeed, it is on the prairies that the old type of cattle industry reached its highest development. The prairies are in reality wet meadows. Their grass cover is due to water relations, most of them being periodically overflowed—conditions that are inimical to pines and palmettos. On the other hand, the period of overflow is too brief to meet the conditions necessary for cypress and other swamp trees. These prairies stretch from the border of the pine woods and palmettos on relatively high ground to permanently wet swamps. The best natural pasturage consists of various species of paspalum and related flat-leaved grasses on the soils fairly moist during a large part of the year; and maiden cane on still moister land, or even in shallow water. Generally speaking, the moisture relations of the more extensive prairies are nearly ideal for continuous pasturage in the varying seasons. There is grave danger in any extensive drainage operations, as palmettos and pines will quickly invade such drained land, and thus destroy the grass.

For improved pasture on these lands, particularly on those reasonably moist, Para grass offers great possibilities. The remarkably rapid growth and high yield of this grass, combined with its palatability and nutritiousness, make it of prime importance in connection with better live stock. Para succeeds well also on the better uplands, but, generally speaking, it is a grass for moist lands. The farther south, the more valuable it is, as after frost it is of little value.

Another grass that is likely to be very valuable on the prairies, and, indeed, on the flatwoods and better uplands, is paspalum dilatatum, native to Argentina. This is perhaps the best of the paspalums, and it is now widespread in the Southern States. Unfortunately, with us the seeds are largely destroyed by a fungus, but good commercial seed is obtainable in quantity from Australia.

Muck Lands.

On the muck lands the problem of pasturage is easy. At least four grasses, namely, Para, Carib, Rhodes and Bermuda, especially Giant Bermuda, yield wonderfully. The enormous area of muck lands in Florida, especially in the Everglades, can, it would seem, be utilized only with the aid of livestock. While there may be some fairly difficult problems to solve in handling live stock on muck soils, especially in the wet season, there can be little doubt that grass and live stock will insure the permanency of these lands. Under continuous cultivation there is a constant shrinkage in muck soils, but with grass and live stock this is nearly, if not quite, counterbalanced.

Carib grass on muck soils is, from limited data, superior to Para grass both in yield and quality. On other types of soil Para will outyield Carib. Rhodes grass does wonderfully on muck soil, and, indeed, on most rich soils. Giant Bermuda is far coarser and more vigorous than ordinary Bermuda. It will succeed wherever ordinary Bermuda will grow, and, in addition, seems much better able to withstand flooding.

Temporary or annual pasture crops are mainly important in connection with swine raising. Various systems of such crops have been devised to furnish successive pastures. Florida has a long list of such crops that can be utilized. Among them are oats, rye, rape, sorghum, peanuts, cow peas, chufas, sweet potatoes, corn and velvet beans. Under certain conditions the cattleman may have to utilize one or more of these crops, but corn and velvet beans is the one that is the most important.

The story of the velvet bean is really one of the romances of agriculture. Introduced into Florida about 1875 from some unknown source, it first attracted attention as a forage about 1890. Until 1914 it was but little grown outside of Florida. In 1915 the crop was certainly less than 1,000,000 acres. In 1916 it had increased to 2,500,000, and in 1917 to about 6,000,000 acres. The explanation of this remarkable increase was the finding of earlier "sports." Three of these appeared independently—one in Alabama, two in Georgia. These early varieties immensely increased the area over which the velvet bean can be grown, so that now it embraces practically all of the cotton belt. These early sports of the old Florida are most grown, but the Chinese velvet bean, introduced by the Department, and the hybrids developed by the Florida Experiment Station, are important. In spite of vigorous search, the native home of the Florida velvet bean yet remains unknown, but it is probably in the Indo-Malayan region of Southern Asia.

The importance of the velvet bean to the live stock industry now developing in the South can scarcely be over-estimated. Grown with corn, it increases the corn crop year after year, and besides furnishes a large amount of nutritious feed to be eaten by the animals when the grass pasture season is over. It reduces greatly the cost of finishing of beef animals for market. This year the velvet bean has been no small factor in helping out the great shortage of foodstuffs, quantities of them having been shipped to Texas. Finally, it has resulted in a new industry for the South, namely, the manufacture of velvet bean meal, which has already won for itself a large demand.

Hay Plants.

The problem of producing hay in Florida is made particularly difficult by frequent rains, except in the fall of the year. The bulk of the hay now produced is from the crab grass that volunteers in cultivated fields. In recent years much Natal hay has been grown for market. Para grass hay is of good quality, and Rhodes grass of very fine quality. Other hays are made from cow peas, cow peas and sorghum mixed, Mexican clover, beggar-weed, oats, millet, etc.

The subject of hay, however, is vital only to the city market. To the live stock man it is of minor importance, as silage furnishes so satisfactory a substitute.

Ensilage Crops.

Corn is, of course, the standard crop for ensilage, and its relative importance in Florida is not far different from that in other States.

Under certain conditions sorghums will yield greater tonnage than corn, and the resulting silage is but slightly inferior.

Florida possesses, in addition, a unique silage plant in Japanese sugar cane. The perennial nature of this plant and its high yielding capacity make it a cheap fodder to grow. It may be utilized as green feed, as silage, as dry fodder, or for pasture. Your own experiment station has published the best information we have on this forage. As a feed for dairy cows there can be no question of its high value, either green or as silage. There still seems to be question, however, as to the relative value of Japanese cane silage as compared with corn silage. In