122 Wisconsin Academy of Sciences, Arts and Letters. 



off the j)oints, jumping under films of this nature, apparently at the glisten- 

 ing air-bubbles which were entangled at the surface; but though the entire 

 range of the fly-book was tried, they persistently refused the hook. In 

 other parts of the lake where there was little scum, the fish, especially 

 pickerel, bit more freely. 



The floating bodies, which at times are as large as pin heads, have once 

 or twice been noticed as a cause of the dispersion of sunlight in a beautiful 

 manner. One calm, bright morning in the fall, a student while rowing, 

 observed a spectrum on the surface of the lake, which took the form of a 

 parabola, with its vertex at his boat. To another person on shore, the spec- 

 trum appeared nearly straight. The apjaearance was of the same nature as 

 the rainbow, and was apparently caused by refraction in the floating dots, 

 which slightly projected from the smooth surface of the water, that cov- 

 ered them with a thin film. A breeze quickly ended it. 



The working of the lakes at Madison corresponds to what is known in 

 parts of Great Britain as the ' ' breaking " of small bodies of water. On the 

 continent, a similar scum is spoken of as water-bloom (Wasserbluthe, 

 Fleur d' eau, Flos aquae). In all these cases, the phenomenon depends 

 uiDon the multiplication of minute algae belonging to tlie group Cyano- 

 phycese, which usually accumulate suddenly and often disappear after a few 

 •days, either temporarily or for the balance of the season. Sometimes the 

 bloom is said to disappear at nightfall, in quiet water, to re ajjpear the next 

 morning. The futihty of fishing at such times is generally recognized in 

 Europe, the common impression being that the fish are then sick. In this 

 connection it may be said, however, that the wholesale death of fish, es- 

 pecially perch, in Fourth Lake, during the summer of 1884, and the death 

 ■of numbers of whitefish every summer has apparently no connection with 

 the water bloom. 



Thuret divided the Cyanophyceae into Chroococcaceae and Nostochineae; 

 the former with their cells not in chains; the latter with their cells placed 

 end to end, forming filaments. The Nostochineae have since been separ- 

 ated into several groups equivalent to the Chroococcaceae, but for our pur- 

 pose the simple division into the two main groups may be retained. 



Of the Chroococcaceae, two are pre-eminent as forming water bloom: — 

 Clathrocystis aeruginosa and Coelosphaerium Kutzingianum. The princi- 

 pal part of the scum first collected at Madison consisted of the former spe- 

 cies (Fig. 8), though the latter (Fig. 7) was also present. 



What have usually been called Anabaena flos aquae and A. circinalis are, 

 in their several forms, the commonest of the Nostochineae that occur under 

 similar circumstances. A small part of the Madison scum of 1882 was 

 composed of sterile plants apparently of A. flos aquae, which, with a httle 

 of a smaller Anabaena, occurred intermingled with the Clathrocystis and 

 Coelosphaerium. These species lasted through the summer, the scum which 

 they formed also containing not infrequent threads of Lyngbya nollei (Fig, 

 ■6), a few stellate plant hairs, and scattering grains of pine poUen. As the 



