OF THE POLLINIUM IN ASCLEPIAS CORNUTI. 77 
later period, be found exhibiting two nuclei in close proximity to each other. In lon- 
gitudinal sections they may be seen to lie in numerous obliquely directed rows, arranged 
one above the other. 
But all the narrow prismatic cells contained in a loculus remain parallel and closely 
appressed together, in close and intimate connexion one with another, so that they 
cannot be separated one from the other without injury and rupture; in the relative 
thickness of their walls, moreover, they present no difference which would enable one to 
assert with any degree of certainty, when this stage has been reached, that any special 
aggregation of cells was the direct derivative of one of the segments of the primitive 
mother cell. The coherent tissue completely filling the cavity of the loculus, and 
bounded by the tapetal membrane, has throughout thoroughly the appearance of a cell- 
mass all of whose cells have been repeatedly bisected in succession by a series of divi- 
sions in two planes only. 
At this stage of their development they correspond exactly to the contents of a single 
loculus in the young anther of Zostera, a genus of Monocotyledons whose mode of 
pollen-formation has been studied in a most masterly manner by Hofmeister *. Indeed 
the earlier stages of Asclepias and those of the last-named genus exhibit an extremely 
close correspondence with one another, the only marked difference between the two 
cases being that in Zostera the anther is quadrilocular. My observations up to this 
point accord at first with the single recorded observation of Schacht, rather than with 
those of Hofmeister, though they commence at a much earlier stage than was noticed by 
either of these writers. Hofmeister regards the pollen as derived, not from a single 
primitive mother cell, as seen in transverse section, but from a group of primitive mother 
cells. Being unable to trace his ** group " of cells back any further, he regards Schacht’s 
statement and figure Т that in Asclepias only a single primary mother cell is formed in 
each anther-lobe as erroneous. The ultimate conclusion reached, if his observations on 
this point be accepted, is, of course, a multicellular archesporium ; while my own results 
distinctly prove that it is unicellular, and that Schacht’s statement really represents the 
true condition of the case. Apart from this, however, I have been able completely to 
eonfirm Hofmeister's researches, in so far as they relate to the pollen-development of 
Asclepias up to the stage at which the most obvious resemblance to that of Zostera is 
exhibited. 
The cell-walls of the primary mother cell and its derivatives by division are thin, and 
always remain so, never being visibly thickened at any subsequent period. In this 
feature they resemble, so far as is known, only Zostera and its near ally Naias 1, while 
they differ in it from the rest of Angiosperms generally. 
At this point, however, the close analogy to, and correspondence with, the type of 
pollen-formation in Zostera ends. In the latter the ae protoplasmic contents 
* Loc. cit. pp. 125-128, plate iii. figs. 4-15 5 ; also Neue “Beiträge, i ii. pp. 643-645 ; also Johannes Grénland, 
* Beiträge zur Kenntniss der Zostera marina,” Bot. Zeit. 1851, pp. 185-192, and plate iv. 
+ Das Microskop, pl. iii. fig. 8; English edition, 1853, p. 105, fig. 21 a. 
i W. Hofmeister, Neue Beiträge zur Kenntniss der Embryobildung der Phanerogamen, part ii. Monokotyledonen, 
pp. 642, 643, plate i. figs. 1-12, 1859. 
N 2 
