McNicol. — On Cavity Parenchyma and Tyloses in Ferns. 409 
points it is not noticeable in transverse view, though in longitudinal view 
the cells are seen to be enlarged and crenulated. Fig. 16 is a section 
of the cavity parenchyma showing the lobed nature of the cells. In the 
Water Ferns, as in the Land Ferns, the cells grow out first into the lumina 
of the vessels by small swellings which enlarge within the passages. 
With regard to the function of the cavity parenchyma thus widely 
distributed in Ferns, Dippel, writing of Osmunda % states that the strands are 
‘ wahrscheinlich zur Aufnahme von Absonderungsprodukten bestimmt und 
den Harz-, Gummi und Milchsaftgangen der Phanerogamen an die Seite 
zu stellenk Terletzski s view differed from this: his opinion was that the 
strands had £ keine andere Funktion als die Geleitzellen, da sie von diesen 
nur in Grosse und Form abweichen, aber im Bau der Wande und in Inhalt 
mit letzteren iibereinstimmen ’ (Terletzski ’ 84 , p. 464). It is very natural 
at first sight to look upon the strands, as did Dippel, as special secretory 
passages, for their appearance is so very unlike the normal arrangement 
of the wood-parenchyma cells, but the view is incorrect. The cells are 
generally, it is true, of an ordinary parenchymatous nature, as Terletzski 
thought, but in a considerable number of Ferns, for example, in Dicksonia 
ant arc tic a and in L ox soma their walls are lignified. Such lignification might 
indicate that the function of the tissue was mechanical ; the spiral vessels 
early become functionless, being broken up by the cells of the xylem 
sheath, and their place taken by a new tissue. Such a tissue would naturally 
be much stronger if the cells were lignified, and to a certain extent the 
tissue may have this function. In Cibotium princeps some of the cells are 
reticulately lignified, whilst others are entirely unlignified. Possibly this 
lignification is to compensate for the disintegration of the spiral vessels, 
the amount of conducting tissue being thereby very slightly lessened, 
but as the lignified cells are not all actually in contact with each other, and 
are often isolated, either singly or in groups, such a theory is hardly 
applicable. In the Equiseta, Strasburger states that the watery sap travels 
along the carinal canals and passes from one canal to another at the nodes 
by means of large pitted tracheids which form a connected mass. Large 
swollen parenchymatous cells near the tracheids frequently project into the 
carinal canals and have something the appearance of cavity parenchyma 
cells, though otherwise in the Equiseta the protoxylem is replaced by 
a canal which is used for water-conduction. But as in the Ferns, the 
ordinary tracheids composing the metaxylem are sufficient to carry the 
stream of water up the plant, it seems likely that the tracheids found 
in the cavity parenchyma of Cibotium are simply for storing water, and 
that in other Ferns where lignification of all the cells occurs, the tissue 
performs the same function. The need for such a storage of water in 
some Ferns is indicated in Todea superba in which Seward and Ford (’ 03 ) 
describe some tracheids of great breadth found here and there in the xylem 
