also has fired' badly in some places. It is noticed that crops on the manured plots are suffering, while the unmanured cotton and tobacco are growing well and are not suffering at all yet. Opinion is divided in the neighborhood as to whether the cotton is in immediate need of rain."

By the 5th of August tobacco had been much injured, and cotton was suffering. On the 1st of September the longest and most severe drought of the season commenced, there being no rain thereafter for nearly six weeks.

The following table gives the mean weekly temperature of the air and soil at different depths at 7 A. M., 1 P. M and 7 P. M., and will be found useful and interesting in connection with the above table of the weekly moisture determinations. More work has been done, and still more is proposed, on this subject of the moisture of the soil, not only with our own soil, but, it is to be hoped, with other typical soils of the State.

ON THE TEMPERATURE OF THE SOIL.

The soil thermometers used by us were described on page 97. The readings were taken at 7 A. M., 1 P. M. and 7 P. M., giving equal intervals of six hours between each reading.

The thermometers were permanently buried to their respective depths in May, and have not since been removed. They were put in the unmanured cotton field where the work just recorded was carried on, the land sloping 21° to the south-east. It was intended to have the cotton growing around the thermometers, and to have the land around them cultivated as if they themselves had been cotton plants. The cotton, for the sake of uniformity, was put in in checks 2x3 feet apart, but with so much passing back and forth to read the instruments, the adjacent plants were injured, so that, practically, the thermometers were never shaded. The soil was kept worked about them as in the rest of the field.

About all we can hope to do in this subject of soil temperature for some time to come is to collect data from which, when a sufficient amount has been obtained and sufficient mathematical knowledge brought to bear on it, the relation of the soil to heat will be established-a fact which will prove of practical value to agriculture, more especially, perhaps, when special branches are considered, as tobacco culture, trucking, grape culture, and the like. The time will come when the tobacco grower, market gardener and vineyardist will study his soil thermometer just as he now does the temperature of the air.

That every soil has a definite relation to heat is undoubted, and unmistakable proof can be seen by a careful study of the statistics gathered here and elsewhere, and of the "soil temperature curves" on the plates at the back of this report, where the mean temperature of the soil, for certain periods of time, is illustrated by diagrams.

In this, as in all scientific investigations, we must have this aim before us constantly-not the mere collection of facts in regard to the variation of this thermometer or that, but we must have a working hypothesis-something to prove or disprove;

and we must collect and apply all our results to this end. Such an hypothesis we have established for ourselves. We hold that there is a definite and important relation between the temperature of the air and of the soil at different depths, that this in a measure is under our control, and that it is one of the factors, if not the most important factor, both of crop production and of crop distribution.

In the diagrams or "curves" at the back of this report can be readily seen the effect of the daily variation of temperature in the soil, which practically ceases to be felt at a depth of twenty-four inches. The annual variation is felt at a very much greater depth, and as it affects the 24-inch thermometer about the same as it does the 3-inch one, the whole figure is caused to rise or fall, but maintaining almost the same relative shape. The effect of these two periodic changes of temperature is to produce waves of heat, which descend into the earth and gradually die away.

Maxwell says (Theory of Heat, page 265): "The length of these waves is proportional to the square root of the periodic times. If we examine the wave at a depth such that the greatest heat occurs when it is coldest at the surface, then the extent of the variation of temperature at this depth is only of its value at the surface. * * * Since the depth of the wave varies as the square-root of the periodic time, the wave length of the annual variation of temperature will be about nineteen times the depth of those of the diurnal variation."

In the rocks of this country it has been determined that, at a depth of about twenty-five feet, the temperature is six months in arrears, being coldest there when warmest on the surface, while at about fifty feet the temperature is twelve months in arrears. end has only a very small fraction of its value at the surface.

To work this subject up fully requires considerable mathematical knowledge and more time for calculation and observation than we have been able to take from our other work up to this time.

We need to know the time when each of the thermometers reach their maximum both for the daily and yearly variations. This has been only partially determined on the Farm for one day,

when hourly readings were taken from 7 A. M. to 7 P. M. On July 28th the maximum occurred approximately as follows: air, 3 P. M.; surface of ground, 1 P. M.; three inches deep, 3 P. M.; six inches deep, 5 P. M., while the 12-inch thermometer was still going up at 7 P. M., and it has frequently been observed that the 24-inch thermometer is warmest at 7 A. M., although its daily variation is very small. It should also be stated that on this date the 12-inch thermometer fell 0.4° between 7 and 9 A. M., which loss was not regained before 11 A. M. Indications of this minimum occurring after 7 A. M. have been noticed frequently in drawing the daily and weekly "curves," and can be very plainly seen in the weeks ending October 8 and October 15, when the soil was very dry and nightly radiation at a maximum. Frequently, after an unusual amount of radiation from the surface, the minimum occurs at 12 inches (apparently) even after 1 P. M., but this seems to occur more often at the New York Experiment Station than with us.

In working over the report of the New York Experiment Station, it has been frequently noticed that the temperature of their soil twelve inches deep is the same or even lower at noon than at the morning reading. This can be seen in the "curves for July in 1883 and 1884, and many other instances bearing on this could be given. One would be inclined to believe at first sight that the temperature had remained constant and had not yet been affected by the daily rise. This was at first attributed to a peculiarity of their soil as compared with our own; but if one carefully examines our curves for July 28, also those for the mean temperature for the weeks ending October 8 and October 16 and other cases, it will be seen that the 12-inch thermometer regularly fell after the morning reading, and it seems probable that by a singular coincidence they have selected the hour (noon) when the temperature at that depth, having passed its minimum (say at 10 A. M.), has again just reached the point on its daily rise, where it was, when observed on its fall, at 7 a. M.

Our soil themometers do not read lower than about 30°F., so we have been unable this winter to determine the time of the annual minimum.