ON THE COI.OKATION OF THE CHINA SEA. 67 



transparent spaces. Tlie red colour predominates in the true 

 China Sea [Nan-Hai), which washes the shores of the south 

 part of China, to the south of the island of Formosa ; whilst 

 the yellow colour predominates to the north of the island, and 

 in the sea specially called the Yelloiv Sea (yHoanf/-Hai). The 

 cause of this phenomenon is unknown. The English who 

 trade in these latitudes attribute it to the spawn of fish — a 

 popular explanation frequently given for all kinds of marine 

 phenomena, and which had already been applied particularly 

 to the coloi'ation of the Red Sea. 



M. Mollien collected a certain quantity of this coloured 

 water, and, on his return to France, he kindly intrusted it to 

 me for microscopical examination. He sent me at the same 

 time the following note of the conditions under which the 

 water had been obtained: — "The sea-water was drawn up, 

 the 14th September last, in 10° N. lat. and 106° E. long. 

 This water was not yellow, as in the canal of Formosa, 

 but red." 



The quantity of the water I examined was very small ; it 

 had deposited a sort of mud of a brown colour which I placed 

 under the microscope. I recognised that this deposit was not 

 formed, as one might at first have supposed, of earthy parti- 

 cles, but that it consisted entirely of an agglomeration t)f 

 minute Algae, almost microscopical and more or less decom- 

 posed. 



These plants presented the appearance of little bundles, 

 which cannot be better described than as resembling packets 

 of cigars, and which resulted from the juxtaposition of a cer- 

 tain number of slender filaments, much longer than broad, of 

 the same diameter throughout, and terminated by rounded 

 extremities. These filaments were probably united by a 

 mucous substance ; but the state of these little plants did not 

 allow me to ascertain this point. They were divided by a 

 great number of transverse partitions into a series of cylin- 

 drical cells, the transverse diameter of wliicli was nearly 

 twice the longitudinal. These cells were slightly constricted 

 in the middle, a sort of indication of their ulterior division. 

 The membrane of the cells was colourless, but the colour 

 might have disappeared in consequence of the incipient de- 

 composition which the plants had undergone. In their interior 

 a certain number of very fine granulations might be observed, 

 which were slightly tinged with yellow. For the rest, the 

 little plants were for the most part much changed • a gi'eat 

 number of filaments were scattered in the fluid as v.ell as a 

 great many isolated cells resulting from the disintegration of 

 other filaments. The study that I made of these little plants, 



F 2 



