sometimes in such dense masses as to colour the sea. Scoresby (Arctic Regions, i, p. 176) noted that in the Greenland Sea the colour of the water was in places nearly grass-green. In the North Atlantic Ocean, in the spring of 1893, I saw bands of brilliant emerald green, like green meadows stretching for miles over the otherwise bright blue sea. And in the South Atlantic the Scotia passed through similar bands of a bright orange colour. These orange bands were fully thirty feet wide, and stretched several miles in length, and Dr. Rudmose Brown found they were composed of a gelatinous scum consisting chiefly of microscopic algæ (Trichodesmium) closely allied to diatoms. It is interesting to note that in this scum were numbers of Portuguese men-of-war, jelly-fish, swimming-bells, and crustaceans, and many other forms of animal life.
I make special mention of these remarkable occurrences because it is quite plain that all these animals were there dependent directly or indirectly on these unicellular algæ; some of the animals were feeding on the algæ themselves, others were preying on those very animals which had become luscious with the good pasture they had fed upon, and these in their turn were devoured by their larger and more rapacious brethren. Why the diatoms were there is a more puzzling question, but there
