96 | BRYOLOGY OF NEW ZEALAND. 
The structure of the leaf in the more robust forms of the New Zealand 
plants is that which Cardot terms “ heterostrosic ’—that is to say, that 
while the leaf-section in the middle and upper parts of the leaf shows a 
single ventral and a single dorsal layer of hyaline cells Capiotte). with 
the chlorocysts saeliany the basal section shows several layers (2-3) of hyalo- 
cysts on both the ventral and dorsal sides of the chlorocysts, Ros edidiag 
to a considerable thickening of the leaf. This thickening, however, does 
not extend across the whole width of the nerve, as along the median line 
the hyalocysts continue to be in two layers only, one ventral and one dorsal 
eae so as to form a neck, so to speak, or isthmus, which may be of varying 
ength ; sometimes only a single cell on the median line of the leaf being 
al: while in other cases 3-5 series of cells on either side of the median 
line are undivided, the narrow neck then occupying a considerable proportion 
of the width of the leaf-section. In this case the — is practically iden- 
tical with that figured by Cardot for L. adunc & Molk. ( 
“Recherches Anatomiques sur les Leucobryées,” pl. iv, Se: 14, a). In the 
smaller forms, notably those which usually pass as L. brach yphyllum ee a 
the thickening of the ventral and dorsal layers of hyalocysts at the lea 
base is very much reduced, and may even be confined to a few cells on 
each side of the median line. This is the state figured by Cardot (op. eit., 
vi, fig. 24, a, b, ce). I have, however, not met wit this extreme state 
Zealand showing 2-3 layers of hyalocysts on the dorsal face at least, over 
a considerable part at least of the section near the base, usually wit 
very short isthmus or neck ; while I have observed every stage of gradation, 
both as to the multiplying of the hyalocyst layers and the length of the 
“isthmus,” up to the most extreme form as in L. aduncum cited above. 
This degree of development is usually parallel to the greater or less size 
of the leaf; and where, as frequently occurs, two forms of leaf are found 
on the same plant, the aiialler leaves will sometimes show the simpler, the 
larger the more highly developed structure. The ma es sts are centric 
above, —. centric or slightly hypercentric towards 
the most striking forms is that named by C. “Midler L. speiro- 
hee aad referred to under that name by Geheeb in Rev. bryol., 1876, 
p. 3, and also by C. Miiller in the Gen. Muse. Fr., but in each case unaccom- 
panied by any description. It has the leaves strikingly seriate in 5 rows, 
and usually more or less faleate at the tips. I fail, however, to find any other 
character, structural or otherwise, to separate it from L. candidum, the 
fruiting characters showing no differences. Geheeb (loc. cit.) writes of 
L. speirostichum, “Selon M. Hampe c’est le L. Teysmannianum.” Of the 
two Javan species with seriate leaves, however, it is not L. Teysmannianum 
so much as L. — Doz. & Molk., with which it should be compared ; 
since L. Teysmannianum, so far as it differs from L. pentastichum, does so 
in the failure of he Syalne leaf-border to reach the apex, while in L. pen- 
tastichum this border is continued, in 2-3 rows of cells at teat. to the 
nde 
| tip, as is the case with L. candidum. I am, in fact, unable to separate 
OL. sper 
this capes a — of Fleischer’s is interesting. Under L. penta- 
stichum (Musci_ . n Buitenzorg, i, 148) he writes: ‘' L. strictifolium 
Broth. in pet aus i Neve Sud- Wale, scheint eine kurz und geradblattrige 
-Abart unserer Art zr Now, I am convinced, from the specimen 
