92 MR. H. M. WARD ON THE STRUCTURE, DEVELOPMENT, 
ceptible, cloudy nucleus-like body appears; but I am by no means convinced that this is 
a distinct nucleus *. Some of the spheres appear orange-red from the first, on account 
of the large quantity of colouring oily matters present. Whether solution and formation 
of these matters, and consequent preponderance of green or red, follow a definite course 
in these earliest stages, I have been unable to decide. 
This spherical mass of protoplasm becomes closely attached to the cuticle of the leaf, 
and soon becomes divided by about two or three division-walls into a few segments of 
equal or nearly equal volume. This primary process of division appears, in typical cases, 
to occur as follows :—three or four pale slit-like tracts or lines appear in the green- or 
orange-coloured groundwork (Pl. XX. & ХХІ. figs. 48 & 50), arranged in a star, and ra- 
diating from the centre; at first each of these pale lines appears isolated in the protoplasm ; 
but their ends become at length united in the centre, and before long also cut the circum- 
ference of the mass. I have failed to prove that these first minute division-walls consist 
of diffluent cellulose; but since the cell-walls produced later consist of this substance, 
it appears highly probable. If the three or four radial slit-like tracts are the expressions 
of so many “ cell-plates "t, as I believe to be the case, the primary process of division of 
this body becomes intelligible. Unfortunately, I have not been able to determine this 
point satisfactorily, or the exact relation of the nucleus-like body to the division-process. 
If we have a nucleus in this early stage, the primary division might be simply one of 
tetrahedral segmentation, one of the tetrads suffering no further division. 
Bethis as it may, the divisions immediately following are simple enough. The triply 
or quadruply segmented body is now a disk, closely flattened to the cuticle of the leaf. 
Each of the segments becomes again bisected, or nearly so, by walls radiating from the 
central to the peripheral regions (figs. 49 & 51) ; and as the new segments increase in size 
and the margin of the disk becomes extended, other radial walls follow, and a regular 
disk of larger size, and consisting of twice as many segments, is produced. After a 
number of radial walls have thus appeared, one or two tangential septa are formed 
across the older cells or segments; and these processes are repeated indefinitely. The 
diagrams (fig. 53) may serve to make these points clearer. 
lt will be evident that the disk, or young thallus, here follows the law of growth 
known as “dichotomous,” though slightly modified j. Each radial row becomes more 
divided, as it lengthens, by tangential walls; and each marginal cell, as it widens, be- 
comes bisected, as a rule, by a radial wall. In accordance with this arrangment and 
mode of growth, each cell is elongated in the same direction as the longitudinal row of 
which it forms a part. It will also be noticed that the radial walls tend (at least in a 
large number of cases) to eut the walls on which they abut at right angles §. If 
every cell grew with equal rapidity and vigour and the divisions followed quite 
* Nuclei are now known to occurin many freshwater Alg hitherto considered to be without them. See Schmitz, 
Bot. Zeit. 1881, p. 32. 
+ Strasburger, * Zellbildung und Zelltheilung. See also a paper by J. T. Cunningham in Quart. Journ. Micr. 
Sci., January 1882. 
t Cf. also Schwendener, * Das Mikroskop, * Die Gesetze der Zellentheilung.’ 
: $ Cf. Sachs, * Anordnung d. Zellen &e. " ; also Pringsheim, in Jahrb. für wiss. Bot. ii. 1; and Rosanoff, “ Recherches 
. sur les Mélobesiées," in Mém. de la Soc. Imp. de Cherbourg, 1866, tom. хи. 
