FILICINEM. 
419 
if the history of its development alone decided the question, and then it would 
offer no analogy to the same structure in Angiopteris. But analogy clearly indicates 
that in Marattia we have not to do with a multicellular sporangium but with a 
sorus, the individual sporangia of which have become united. Like those of 
Aitgi'opteri's, each of these sporangia opens by a longitudinal slit upon its inner 
surface. It is of but little importance for this interpretation that the apparently 
multicellular sporangium, which we regard as a coalesced sorus, is borne in 
Eupodium {Mar. Kaulfussii) on a stalk of considerable height, for the sorus in 
many of the true Ferns {Cyathea^ Thyrsopteris) is also stalked. It can scarcely 
be doubted therefore that the multicellular fructification of Maratiia is a coalesced 
sorus, and the same holds good also for Kaulfussia and DaficBa. In Kaulfussia 
the sporangia of a sorus (from eight to twenty in number) are arranged in a circle 
and are united to form a many-chambered ring. Each opens on its inner side 
by a longitudinal slit. This arrangement is even more striking in Daiicea, where 
the united sporangia form two long rows covering the vein bearing them throughout 
its whole length, and where each chamber (sporangium) opens at its apex. The 
sorus is usually surrounded by flattened lobed hairs forming a kind of indiisimn, 
which, in DancBa^ appears like a kind of cup in which the sorus lies. Luerssen's 
argument that these outgrowths of the epidermis are not to be regarded as an 
indusium because they occur elsewhere upon the leaves and are therefore merely 
hairs, is not valid, for the indusium of the true Ferns is a hair-like outgrowth, and 
must be regarded as a trichome. As in the Ferns, so in the Marattiaceae, the 
indusium does not occur in all species. 
The development of the sori has been studied by Luerssen and by Goebel 
in Maratiia^ and by these observers and by Tschistiakoff in Angiopteris. In both 
cases the placenta arises as a cushion-like protuberance from the fertile vein of 
the epidermis and the subjacent tissue. In Angiopteris two separate rows of 
papillae make their appearance upon the receptacle, each of which consists from 
the first of a group of cells derived from a group of the superficial cells of the 
placenta. Each papilla becomes one of the free sporangia of the sorus. In 
very young sporangia Tschistiakoff was able to detect an internal cell (arche- 
sporium) surrounded by two or three layers of cells which gave rise by repeated 
division to a group of spore-mother- cells. In Marattia two parallel swellings 
appear on the placenta, which soon become separated by a deep and narrow fold. 
In each of these swellings a row of cell-groups, the mother-cells of the spores, 
are differentiated, which have been formed by the division of the archesporium. 
Each of these groups corresponds to a sporangium, the walls of adjacent sporangia 
coalescing from the first. The inner surfaces of the two parallel swellings approach 
each other more and more closely as development proceeds, but they separate 
widely when the spores are ripe, so that the multilocular fructification splits longi- 
tudinally into two halves, and the loculi of each half open by vertical slits upon 
their inner surfaces. 
The development of the spores, four from each mother-cell, differs but litde 
from that of the Ophioglosseae and the Ferns. It is important to note that in 
the Marattiacese the wall of the mature sporangium consists of several layers of 
cells, whereas in the Ferns it consists only of one. 
E e 2 
