so that light of the intensity 50, acting during time t, produces the same blackening effect as light of intensity t acting during the time 50. According to this method the chemical action of the total daylight was determined for Manchester, England, many times a day during 1864, and the total daily chemical intensity has been computed for the year August, 1863, to September, 1864. Very large changes in chemical intensity occur when the sky is cloudless and unchanged as far as the eye can perceive. The total intensity for an apparently cloudless day varies from 3.3 for December 21, 1863, to 119, June 22, 1864. This last number, compared with the figure 50.9 for June 20, and 26.6 for June 28, shows the enormous variations that take place in the chemical rays that reach the observer at Manchester on cloudless days. This variation is undoubtedly due in part to smoke and moisture, but possibly other unknown influences are also at work.

In 1867 H. E. Roscoe communicated to the Royal Society the results of work done by his method at Kew, England, in 1865, 1866, and 1867; at Heidelberg, 1862 and 1863, and at Para, Brazil, 1866. The general results are that the chemical intensity attains its maximum at noon and not, like the temperature, at some time after noon. Everywhere the intensity increases from hour to hour with the altitude of the sun, and is very closely proportional to it even when the sky is partially clouded, but of course the rate of increase varies with the season, the amount of cloud, and the degree of atmospheric opalescence. The total chemical intensity for each month, as determined from numerous observations, is as follows for Kew:

Total photochemical intensity of direct and diffuse light (Roscoe).

Roscoe compares these figures with the cloudiness, and finds that the ratio between cloudiness, expressed on a scale of 10, and the chemical intensity is as 1 to 5 in some months and as 1 to in others. A similar irregularity of ratio is found when he considers the absolute moisture in the atmosphere; whence he concludes that the variations in chemical intensity, as between the spring and autumn, are not perfectly explained by either of these factors. He finds the high autumnal and low vernal intensity fairly well explained as due to the transparency or opalescence produced by finely divided solid particles or dust.

Passing from Kew to Para, it appears that the chemical action of total daylight during the month of April, 1866, at Para was 6.6 times as great as at Kew.

In order to obtain data for a clearer atmosphere, Roscoe and Thorpe conducted observations in 1867 near Lisbon, Portugal, and published their results in a memoir of 1870, where they have given the relation between the sun's altitude and the chemical intensity. The intensity is the same for hours that are equidistant from apparent noon. The relative intensity of direct sunlight, reflected sky light, and total insolation is shown for different altitudes at Lisbon by the following table:

Intensity of insolation at Lisbon for clear skies.

In general, the total intensity is directly proportional to the number of degrees of altitude. For altitudes between 18° and 35° the intensity on a plane perpendicular to the incident rays is about the same as the intensity of total sky light on a horizontal plane. The intensity of direct sunlight on a horizontal plane is equal to the intensity of total sky light on a horizontal plane when the sun's altitude is about 45°. At all altitudes of the sun below 21° the chemical action of diffuse daylight exceeds that of direct sunlight.

In their memoir of 1871 Roscoe and Thorpe determined the amount of chemical action for total sky light of a cloudy sky during totality of the solar eclipse, and found it much less than 0.003, and therefore not measurable. They found the total chemical action of the direct sunlight to be strictly proportional to the visible area of the portion of the solar disk up to a certain point in the obscuration, after which the influence of sky light is inappreciable. For altitudes below 50° at Catania, Sicily, as elsewhere, the amount of chemical action effected by diffuse daylight on a horizontal surface is greater than that exerted by the direct sunlight. At altitudes less than 10° direct sunlight is almost completely robbed of its chemically active rays.

PHOTOGRAPHIC INTENSITY OF SUNSHINE.

A photographic method of determining the brightness of sunshineor sky light is very desirable as supplementing the thermometric methods. It is as erroneous to assume that all radiation that falls upon a black-bulb thermometer is absorbed by it and converted into heat and measured by the expansion of the mercury as it is to assume that all the radiation that falls on a photographic film is absorbed by it and is represented by the chemical changes that take place in the film. Equally erroneous would it be to assume that all the radiation that enters the eye is represented by the impression of brightness conveyed by the retina to the brain. In order to measure in absolute units the total energy radiated from the sun, we need a proper summation of the thermal, visual, and photographic work done by the radiation. If we wish to determine only the intensity of that part of the radiation that does the work in which agriculture is chiefly interested we should consider only the heating effects of the radiation and the special chemical effects manifested in the action of sunlight upon chlorophyll.

The action of the sunlight upon the chlorides and bromides of silver, as in ordinary photographic processes, may not be an exact measure of its action upon the leaves of plants. Some other chemicals may be more appropriate for use at agricultural experiment stations, but the photographic methods perfected by Profs. H. W. Vogel and L. Weber are worthy of trial as a first step in the right direction. These processes give us the relative intensity of the radiations that belong to the blue end of the spectrum, with only a small admixture of the influence of green and yellow rays.

During the year 1890, as the result of a numerous series of observations at Kiel, Prof. L. Weber found that the reddish light of the spectrum on dark winter days has only about 500 times greater intensity than the quantity of light from a normal candle at a distance of 1 meter, when measured by their relative effects on a photographic plate, while at the same time the photographic intensity of the green light of the spectrum was four times as much. On bright summer days the intensity of the red light was 50,000 times that of the candle at 1 meter, while the intensity of the green light was about 200,000, or about 4 times as much in summer as in winter. The intensity of the blue light in the solar spectrum was about 25 times that of the red light, which ratio varied a little with the kind and amount of cloud. In all this photographic work a very sensitive silver bromide paper was used; so that these results, strictly speaking, relate only to the variations in the intensity of those special rays that affect this chemical. But these variations will be nearly parallel to the diurnal and annual variations of the rays that affect the growth of plants.

Further details of Weber's results are given in the German periodical, Photographische Mitteilungen, edited by Professor Vogel, at Berlin.

It is worth while to call attention to the fact that during the long twilights of northern latitudes in midsummer plants receive an appreciable quantity of the blue radiations from the sky, while receiving little or nothing of the red, or heat, rays.

MARCHAND'S SELF-REGISTERING CHEMICAL ACTINOMETER.

66

A convenient form of registering actinometer is that devised by Marchand (1875), which he at first called "photantitupimeter," but which name he afterwards contracted and modified to phantupimeter." This consists of a vertical graduated tube, closed at the upper end, into which there can escape and be measured the carbonic acid gas given off by the decomposition of a mixture of solutions of perchloride of iron and oxalic acid. By the action of sunshine on this mixture, carbonic acid gas is slowly disengaged, and by its accumulation in the measuring tube gives us apparently a means of determining the sum total of the influences of the sun during any period. This apparatus was diligently employed for many years by Marchand at Fecamp, near Havre, and has afforded him many interesting results.

COMPARISON OF MARCHAND'S AND MARIE DAVY'S RESULTS.

Radau (1877), in his work on Light and Climate, states that the results given by different methods of measurement of sunshine appear to differ largely among themselves, but yet there is a certain similarity in the figures. The accompanying table shows the results of observations by Marchand's chemical method and by Marié-Davy's thermometric method, or conjugate thermometers, which latter, on account of its convenience, has been widely adopted.

If the atmosphere were not so very different at these two localities,

we could have hoped to use the monthly ratios of these numbers for reducing similar series elsewhere to a common standard.

VIOLLE'S CONJUGATE BULBS.

The refined methods for measuring solar radiation adopted by Violle (1879) in his absolute actinometry can hardly be utilized in agricultural investigations owing to the labor of using the apparatus. But the continuous register obtained by him by means of thermoelectric apparatus is an important improvement in the methods available for comparing climates. On the other hand, Violle has suggested a modification of the conjugate thermometers which he calls his "conjugate bulbs," which is worthy of consideration, although far from being as sensitive as Marié-Davy's apparatus. These bulbs are made of thin copper, one of them blackened and the other gilded on the outside; the interiors are blackened, and the thermometer bulbs within them are also blackened. This apparatus has an apparent advantage over Marié-Davy's, in that the sunlight is not required to pass through glass before striking the thermometer. It would appear likely that with smaller bulbs (Violle uses 1 decimeter in diameter) and with more sensitive thermometers Violle's method might give better results and be worthy of recommendation to agricultural investigators. The results given by his apparatus have need to be reduced by some method based on the considerations indicated by Ferrel (1891).

BELLANI'S RADIOMETER OR VAPORIZATION ACTINOMETER.

Among the many devices invented for the purpose of obtaining, at least approximately, the sum total of the effect of sunshine received during any day by a given plant is one that has been used for a few years at the Montsouris Observatory, and is a modification of an apparatus originally devised by the Italian physicist, Angelo Bellani, which is thus described by Descroix (p. 128, Annuaire de Montsouris, 1887; see also the Annuaire for 1888, p. 206, where it is called the lucimeter, although it does not measure light properly so called).

The vaporization actinometer or the Bellani radiometer as modified at Montsouris consists of a bulb of blue glass A of about 60 mm. in diameter, inclosed within a larger bulb B of colorless glass. The space between the two bulbs is a vacuum. A is two-thirds filled

with a volatile liquid and the space above it contains only its vapor, which passes through a curved tube down into a large bulb C, of clear glass, and thence down into the vertical tube D, which is graduated, and where the condensing liquid can be measured.

Under the action of the radiation from the sun and the sky the blue bulb A is warmed more than the bulb B; a distillation takes place from A and the condensed liquid is collected in the graduated tube D, where its volume is measured. This condensation in D is a source of heat, while the vaporization in A is a source of cold. The heat given off by condensation must equal that consumed in evaporation, and is drawn off from the apparatus by the action of the cool

2667-05 M- -7