contain a considerable amount of these bicarbonates and soft-water lakes have little or none. In hard-water lakes it is found that during the growing season, when algæ are active, the upper water contains no free carbon dioxide, but is, on the contrary, alkaline, when tested with phenophthalein as an indicator. This alkalinity comes from the fact that one molecule of carbon dioxide has been withdrawn from part of the bicarbonates, converting them into carbonates. It appears that the algæ are able to effect this reduction and that they can obtain their supply of carbon from the carbon dioxide of the bicarbonates dissolved in the water. This fact introduces a wholly new feature into the story of the food supply of the plants. It provides a chemical carrier for the carbon dioxide which may carry this gas somewhat as the hemoglobin carries oxygen in the blood. All carbon dioxide set free in this alkaline water as the result of decomposition or other processes, will be taken up immediately by the carbonates. Thus if plants are not at hand to utilize the carbon dioxide at once, it is not lost but kept until it is needed. So in the night, the lake is able to retain all the carbon dioxide set free and which the plants do not use at that time.
Such alkaline water has also a great advantage in absorbing carbon dioxide from the air. It presents for absorption, not merely the relatively weak and slow powers of the water for dissolving the gas, but the eager and vigorous powers of chemical affinity. And until these alkaline carbonates are saturated, no free carbon dioxide will appear in the water to diminish the rapidity of absorption from the air. Thus hard-water lakes have an advantage over soft-water lakes in the matter of securing plant food, and in fact the population of soft-water lakes is smaller than that of lakes of the other type.
It is worth while to devote a few words to gaseous products of decomposition other than carbon dioxide. So long as the bottom water contains an abundance of oxygen no other gas than carbon dioxide is produced in appreciable quantities. But as the oxygen becomes greatly reduced or wholly disappears, decomposition continues in new forms and under these conditions of anaerobic fermentation other gases may be developed in considerable amounts. It is apparently true that carbon monoxide may be present in the lower water of lakes in appreciable quantity, and it is certain that marsh gas is developed in large volumes in lakes where the amount of fermentable material is large and where the oxygen disappears from the lower water early in the season. These gases first appear near the bottom, where decomposition is going on most actively and where the oxygen first disappears. In many lakes they are found only in small quantities and close to the bottom, but in proportion as the amount of decomposable matter increases, they are found at considerable distances from the bottom, and in certain lakes all of the water below the thermocline may contain
