The water colors of twenty ponds which were selected at random among the eighty ponds were measured by the new method mentioned already at every fifteen days. The mean of chromatic coefficient (x, y), brightness (Y), dominant wave-length (λD), and excitation purity (Pe) of the water colors of these ponds were calculated at every time in order to catch the tendency of seasonal change of water color. The mean of the chromatic coefficient is plotted on the C. I. E. chromaticity diagram (Fig. 1). The mean of the other values are shown in Figs. 2, 3, and 4, by which we can see the tendency of the seasonal change of algal population in the confidence limits of 95 per cent. The results obtained are as follows :
(1) The locus of point x, y, runs from April to November in nearly parallel with the spectralocus (Fig. 1). It may therefore be said that the change of the water color during this period is a change of the hue, not of the brightness and chroma.
The value of the dominant wave-length decreases gradually from 577 mμ during the period from April to July, and at the same time water colors change from yellowish brown to yellowish green through olive (Fig. 2), the value of the dominant wave-length increasing in the opposite way.
(2) The locus of point x, y, runs from November to March in nearly parallel with the isodominant wave-length's line (Fig. 1). It seems therefore that the change of the water color during this period is principally a change of both the brightness and chroma, not one of the hues of the water color ; and at the same time water colors change quickly from yellowish green to dark olive during the period from November to December, returning to yellowish green or pale olive till March (Figs. 1, 3, and 4).
(3) The dominant wave-lengths were distributed between 565 mμ and 579 mμ, mostly between 572 mμ and 574 mμ during the year ; accordingly, the water colors were mostly olive during the year. The value reached its maximum in April and at its minimum in July (Fig. 2).
The brightness is distributed between 7. 0 and 12. 0 per cent and the excitation purity is distributed between 31.0 and 44.0 per cent. The curve of the brightness is considerably similar to that of the excitation purity (Figs. 3, 4).
I have reported the same facts as a simple equation which denotes a positive correlation between the two values.
(4) In the seasonal change of confidence intervals in the means of the three values of water color, the values in August, September, and October, as well as in February and March are not so great as the values in the other months (Figs. 2, 3, and 4).
From these facts it is understood that the water colors in both summer and winter resemble one another and are not rich in varieties.
From the seasonal change of water colors, it will be possible to classify them into two categories, i. e. type A and type B, whose characteristics are as follows :
Type A : - In July the values of brightness and excitation purity of water color become low and locate point x, y, approaching the Illuminant C on the chromatological diagrams ; the water colors in this period show dark olive. Such water colors are ordinary found in winter. Therefore, in the ponds of this type, dark olive coloration of water occurs twice during the year, in June and December (Fig. 5).
Type B : - The low value of brightness, excition purity, and dark olive coloration occur only once during the year, i. e. in December (Fig. 6).
View full abstract