CELL-FORMATION. 121 



happened, that the young free cells should have been seen, 

 in some when the parent-cells were larger, and filled 

 with granular contents (as in many hairs when their de- 

 velopment is beginning, e. g. of Tradescantia) ; in some 

 when the parent-cells, of considerable size, contained a 

 homogeneous and tolerably transparent fluid (as in the 

 young rind and bark of many plants) ; in others when 

 the parent-cells were tolerably large, and more or less 

 filled with coloured, homogeneous or granular contents 

 (as in Caltithamniacea). But in no case could I see the 

 young free cells inside the parent-cells. The formation 

 of the cells of tissues in the higher Cryptogamia and the 

 Phanerogamia, displayed to me nothing but either merely 

 a dividing wall, or two nuclei and then a wall between 

 them, or a larger nucleus which divided into two nuclei, 

 and then a septum formed between these. 



In consequence of these researches I have come to the 

 conclusion, that all vegetative cell-formation is parietal. 

 This conclusion is supported on the one hand by the facts 

 observed, on the other by analogy. Among the facts, 

 some do not contribute to the proof; others make the 

 parietal cell-formation probable ; a few can be explained 

 by it alone. In Griffitksia corallina* (pi. Ill, fig. 9), the 

 parent-cells are very large, as much as * 080 of a line long 

 and more, and sometimes of almost half that breadth. 

 The granular contents lie upon the wall ; the interior is 

 filled with clear, colourless fluid. The cell divides into 

 two unequal cells, at the apex, by a cross wall (fig. 10) : 

 an upper cell, small, disc-shaped, and wholly filled 

 with granular contents (a), and a large lower cell (3), re- 

 sembling the parent-cell in all its parts. If the two 

 secondary cells were formed free, the lower, especially, 

 must be seen during its development, and changes in the 

 parietal firm contents be perceived, when this became 

 dissolved in the parent-cell and reorganized in the secon- 



* Griffithsia belongs to the Floridete ; the Florideee, however, have not a 

 natural relationship to the Algse, but with the Hepaticae and Mosses. They 

 also, like these, possess parietal nuclei in the cells. 



