1 8 The Structure of Dictyosph&ria. 



own specimens preserved in spirit from Grenada. The Ceylon material 

 corresponds with the descriptions of Harvey and Agardh in respect of the 

 hollow interior, while my material shows a solid mass of cells in the young 

 state, the interior cells being less firmly united than the peripheral cells, 

 but ultimately becoming hollow in the older specimens. The first observ- 

 ation was directed to discover the nature of the mysterious discs. These 

 proved to be tenacula emitted from and attached to the cells, and binding 

 the mass together. An inspection of figs. 2b, c, 3, will show that these tenacula 

 differ in no essential respect from those already described by Mr. Boodle 

 and myself in Stmvea* Spongocladia,-\ Microdictyon,% Boodlea,\ &c. 

 Where the attachment of the cells is intimate, as in the peripheral cells, 

 the tenacula are short, and in the internal cells, which are more loosely 

 compacted, the tenacula are produced at the ends of filaments (fig. 3) 

 in the familiar manner. The figures exhibit the mode of attachment so 

 well, that a detailed description is unnecessary. After reaching a certain 

 stage of development the tuberous bodies of D. favnlosa burst, and the 

 lobes appear to go on growing irregularly, giving the mature plant a very 

 different appearance from what it presented in its earlier stages. 



With regard to the cell division into four, some of the Ferguson 

 specimens (fig. i c, upper part) show what I take to be an appearance of 

 the kind described by Harvey and Agardh, but I am not altogether con- 

 vinced that the explanation of the matter is so simple. On this point I 

 hesitate to put forward an alternative view for the present. The processes 

 projecting into the interior of the cells (fig. 4) are accurately described by 

 Agardh, and, like him, I fail entirely to apprehend their significance. 



We have here, then, in Dictyosphcsria favnlosa one of the simplest forms 

 of valonioid organism, and one, moreover, of some interest from the wider 

 point of view of thallus-formation. Struvea, Boodlea, and Microdictyon 

 present to us examples of a branching reticulate thallus of this type, the 

 component members of which are held together by tenacula ; other 

 Siphonocladacece, like Spongocladia and Udotea, show us cases of strands 

 of filaments or fronds similarly connected. Valonia itself, the simplest of 

 these forms, branches, though unpossessed of tenacula for the junction of 

 its members. In Dictyosphcsria favnlosa we have simply an aggregate of 

 similar cells not forming a definite frond, but cohering in an unbranched 

 mass, this colony of units being held together solely by tenacula. It is 

 certainly an extraordinary form of tissue -formation, and represents, it 

 appears to me, the most reduced form of the siphoneous thallus, using the 

 term in the wide Agardhian sense. 



* Annals of Botany, vol. ii., No. vii. t Ibid., vol. ii., No. vi. 



% Journ. Linn. Soc. Bot., vol. xxv., p. 243. Ibid. 



