Jensen has, however, shown that bunt in wheat and smut in oats or barley or rye can be almost wholly prevented by washing the seed before sowing, in water whose temperature is not lower than 130° F. nor higher than 135° F. The sacks to receive the seeds should also be disinfected. Professor Kellerman shows that if the seeds are previously soaked in cold water for eight hours the hot-water wash may have a temperature of 124° to 128°. I infer that the spores of the smut, having been by the winds blown over the field in the ripening period, have stuck to the grains from that time on to the next sowing season. (Agr. Sci., Vol. IV, p. 100.)

THUNDERSTORMS AND OZONE.

A. L. Treadwell seems to have shown that the souring of milk during thunderstorms can not be attributed to any formation of ozone, and is more likely to be due to the fact that the bacteria causing this souring multiply with unusual rapidity during the warm sultry weather that precedes and accompanies thunderstorms. (Agr. Sci., Vol. V, p. 108.)

PRUNING VERSUS CLIMATE.

Kraus (1886) in some experiments on pruning hop vines shows first that those that were not pruned had an advantage in the early growth, especially in the cold and wet of June, 1886, in Germany, but in consequence of this precocity the early ones suffered from frost. Those that were early pruned surpassed them in the harvest.

Those that were pruned late gave the smallest harvest, but of the highest quality, the leaves remaining a beautiful green up to the harvest time, while those that were not pruned or those that were late pruned turned dark and soon yellowed.

This explains why for a long time it has been impossible to define exactly the climate that is best for the cultivation of hops, since it is now evident that changes in the pruning, harmonizing with peculiarities of weather or locality, have so great an influence upon the successful cultivation. (See Wollny, X, p. 236.)

WHEAT, TEMPERATURE, AND RAIN IN ENGLAND.

The wheat harvest of England has been studied by an anonymous writer. (Nature, 1891, vol. 43, p. 569.) I do not know the authorities for his statements as to the character of the harvests from year to year, but reproduce in the following tables the figures given by him as to the general character of the wheat harvests for each year and the corresponding mean temperatures and total rainfall for the months of June, July, and August as observed at the Royal Observatory, at Greenwich. Certain deductions are given by him as to the connection between the harvests and these items of the weather, but a more careful study of the figures convinces me that taken as they stand no infer

ence can be safely drawn from them which will endure the test of critical examination. Any small selection of years may be made which will seem to support some suggested relation between temperature, rainfall, and crop, but other years will be found to contradict this. In a general way good crops result from hot and dry summers and bad harvests depend upon the large rainfalls rather than on the low temperatures. I have added the column of departures and have computed the probable errors of the averages, the study of which shows that the temperatures of the good harvest seasons are not sufficiently above those of the poor harvest seasons to justify the conclusion that warm seasons are intimately connected with good harvests. If, however, we go into more detail and examine all of the fifty-three years from 1816 to 1888, inclusive, and arrange them by the character of the harvests, we find innumerable contradictions. The study of the rainfall with its probable errors, or rather its probable variability, shows a somewhat stronger argument in favor of the idea that large rainfalls accompany poor harvests, and yet here again the contradictions are too numerous to allow us to suppose that this simple statement expresses exactly any law of nature. Thus the largest rainfall of 1888 and the small rainfall of 1886 both contradict this law. In the notes a few statements are made by the author as to special occurrences which seem to him to explain these anomalous cases, and by hunting through the records a few more notes might have been added so that after leaving out the anomalous cases one might say that the remainder accords well with the idea that dry hot summers give large crops and that heavy rains give poor crops. In general, however, it seems more proper to conclude that we are far from having attained the expression or formula connecting the crops and the weather, and that even if we knew this it would be improper to study the crops of England with reference to the temperature and rainfall at Greenwich, or, indeed, any other single station.

English wheat harvests and Greenwich weather.

[Weather in June, July, and August.]

I. SUPERIOR WHEAT HARVESTS.

English wheat harvests and Greenwich weather--Continued.

I. SUPERIOR WHEAT HARVESTS-Continued.

The winter was very mild; the spring very dry.

The winter and early spring were very cold; May was very dry, with much sunshine. d Frost occurred at blooming time.

eThe spring was cold.

The winter and early spring were very cold; May was very wet.

SUGAR CROP AND RAIN IN BARBADOS.

Sir R. W. Rawson, as governor of the British colonies at Barbados, published (1874) a colonial report, printed by the house of assembly, giving an elaborate study of the dependence of the canesugar crop upon the monthly and annual rainfall. Barbados offers an exceptional opportunity for such study, since the cane is the only staple and is nearly all exported, so that the records of the crop are accessible in the customs' returns. Moreover, the number of rainfall records averaged more than 1 to a square mile, being 178 for the whole island and for a period of about twenty-five years, this remarkable system of observations being due largely to the labors of Dr. R. Bowie Walcott, who still resides in the parish of St. Joseph, and was, in May, 1890, on the occasion of my recent visit to him, still active in collecting rainfall data. To his devotion and Governor Rawson's assistance we owe this unique study of rainfall and sugar crop. It is impossible for me at present to do more than give the accompanying Tables I, II, and III of monthly rainfalls and annual crops. The crops, as given in Tables II and III, in hogsheads, are credited to the years in which they passed through the custom-house. The cane is usually gathered and the sugar and molasses shipped between January and May; after the latter date the fields are newly planted and in eighteen months are again ready for cutting, so that the crop of any year has been grown under the influence of the rain of the preceding year and the latter half of the year preceding that. In the second table I give the dates of the first shipment of sugar each year, thus showing whether the crop was gathered early or late, and also the general character of the crop as credited to that year.

Table III illustrates Governor Rawson's conclusion that the crop of any year is influenced only in a slight degree by the rainfall of that year, but depends upon the rainfall of the preceding year. Thus it is arranged according to the quantity of rainfall, and the crop of the following year is compared with the rain of the current year; the wet years are followed by large crops the next year, while the dry years are followed by small crops; the increase being 10 per cent after a wet year and the decrease being 12 per cent after a dry year. The general development of the sugar plant is illustrated in the following extract (see p. 33, Rawson's Report):

The influence of the rainfall in particular months and seasons upon the coming crop is generally felt and admitted, but not known with any certainty. It is believed, writes an experienced agriculturist, that any marked excess of rain during the first six months of the year is injurious both to the crop that is being reaped and to that which is to follow. The cane plant during the early stages of its growth is very hardy and requires but little moisture; the small

early shoots are hard and fibrous, and very different from the large succulent shoots which are afterwards produced and which lengthen into the juicy reed whence the crop is made. In ordinary and favorable years, with light showers during the first six months, the young canes make no marked progress, but the roots are increasing in length and strength, and in the months of July and August the plant begins to sucker, as it is called, and to put out the shoots which form the canes, but these make no great progress in length before the end of August and in September and October, when the rains usually come to their aid at the critical time. They then grow with extreme rapidity, are extremely tender and succulent, and a short spell of dry weather at that time usually does serious mischief. If, however, the first six months of the year are wet, and the young canes are excited to an abnormal rapidity of growth, they are liable to be seriously affected by any interval of dry weather in the middle of the year. Moreover, rainy weather in the reaping season retards the manufacture, and, especially in the black soils which contain an excess of iron variously combined, causes a great loss from the rotting of the canes at the roots.

An illustration of this is afforded by the rainfall and crops of 1860 and the two following years. 1860 was a model year; the rain fell at the right time, and in exactly the average quantity, 57.91 inches, of which 12.46 fell during the first six months. The crop of 1861 would undoubtedly have reached 55,000 hogsheads but for the wet reaping season of that year, in which the rainfall of the first six months was 31.93 inches-6.35 in April, 8.01 in May, and 8.01 in June. The consequence was that the crop only reached 49,745 hogsheads, and although so much rain fell throughout the year (73.82 inches), the following crop of 1862 was only 46,120 hogsheads.

In the same manner the heavy rainfall of 1855 (77.31 inches, of which 30.68 fell in the first six months) was followed in 1856 by only a moderate crop (43,077 hogsheads), although the reaping season of that year was most favorable. The result, however, is by

no means constant.

The sugar-crop records go back to the year 1806, but the returns are only interesting since 1847, which was the first in which the crop recovered from the effects of emancipation in 1839. Since 1847 there has been a steady increase until the crop has attained nearly twice what it was before emancipation. There has also been a slow increase in acreage of canebrake; the size of the hogsheads has been gradually increasing since 1806; there has been a decided increase in the usage of guanos and other foreign manures; there has also been a very decided improvement in the machinery and processes for crushing the cane and manufacturing the sugar."

a Although Governor Rawson was evidently conscious of these progressive changes, and in fact, mentions most of them, yet he does not approximately eliminate their effects by taking the difference between the individual crops and a progressively increasing ideal normal, but takes the difference between the simple average and the individual years; his results, therefore, need to be computed and all the data for this purpose are given in the tables herewith.-C. A.