Ch. XXII.] MIOEOSOOPIC APPBAEANOE. 383 



two occasions did I ever observe the sea discoloured by red 

 matter — once by myriads of minute red Crustacea, in the 

 Indian Ocean, and again by a dense mass of red gelatinous 

 worms in Formosa Channel — but never by Trichodesmium. 

 The characters presented under the microscope by the 

 specimens I first obtained in the Indian Ocean have been 

 already alluded to ; and although the ultimate elements of 

 the various specimens were the same, I met with two dis- 

 tinct forms of Trichodesmium, one on either side of the 

 Malacca peninsula. That on the west side was in the form 

 of a miniature sheaf of corn, while that of the China Sea was 

 a cylindrical bundle of fibres, more or less pointed at one 

 end, but obliquely truncated at the other. This was also 

 the form it assumed on the other occasions on which I 

 examined it. Both these, however, consisted of bundles of 

 cellular fibres of the same character. The bundles were 

 cream-coloured and opaque, and a lens showed that the 

 ends were fimbriated, owing to the component fibres being 

 loose at their extremities. With slight compression these 

 fibres were seen to be cylindrical filaments of unequal length, 

 combined together and interlacing one another, forming an 

 intricate net-work, which resembled unfinished basket-work 

 with the long ends of the osiers sticking straight out. Each 

 filament was long, symmetrical, and unbranched, with a 

 rounded extremity, and even, hair-like outline — divided by 

 transverse septa into rectangular cells haK as long as broad ; 

 and each cell contained some grains of chlorophyll in the 

 centre, which rendered it opaque. Continued pressm-e, how- 

 ever, discharged part of this substance, rendering the ceU- 

 waUs distinct, and ultimately the filament broke up into its 

 component cells, which presented various facets to the eye — 

 some round and some rectangular, proving its confervoid 



