198 THE PHENOMENON OF 



higher colour than the surrounding green mass, which 

 exhibits the deepest colour immediately around them. I 

 could not observe how these first globules, or vesicles, 

 originated in the green mass ; the first of them displayed 

 itself immediately after the gonidia uniting to form a net 

 had passed into a condition of rest, before the cavity of 

 the cell had become excavated, and with each succeeding 

 day appeared new ones, which did not seem to have been 

 formed by division of the first, but to have an inde- 

 pendent origin. Even by the second day these vesicles 

 were found surrounded by a more or less evident border, 

 originally green, and with a somewhat sinuous or denti- 

 culated outline, but soon assuming the shape of an 

 accurately defined, exactly circular envelope. The larger 

 globule originating in this way was from ^th to ^gth 

 millim. in diameter, and did not subsequently increase in 

 size; it exhibited a yellowish or light yellowish-green 

 colour, the nucleus (the globule first formed) always ap- 

 pearing somewhat lighter than the envelope. On applying 

 tincture of iodine, these globules assumed a violet colour, 

 the nucleus seeming to me lighter, more of a wine-red ; 

 after a previous bleaching, by keeping a longish time in 

 alcohol, the envelope was coloured darker, the nucleus 

 paler blue-gray ; by a stronger action of iodine, so dark 

 that the colour was no longer distinguishable. On the 

 application of solution of caustic potash, the envelopes 

 swelled up to about three or four times the diameter, 

 and finally vanished entirely, while the nucleus remained 

 unaltered.* No lamination could be seen in the swelling 

 coats. Thus it cannot be doubted that the envelope, 

 originally infiltrated with chlorophyll, was composed, in 

 its fully- developed condition, of starch, while the starchy- 

 nature of the nucleus remains very doubtful. The 

 described mode of formation of the starch-granules of 

 Hydrodictyon appears to tally better with the theory of 

 Fritsche arid Schleiden, that the concentrically striped 



* Whether or not the nucleus would dissolve by longer action of potash 

 requires a repetition of the investigation. 



