ON ANOMOCLADA. 231 



renifonni-orbiculata, valde concava margineque aatico ])roecipue 

 late incurva, hast antica longe semini decurrentia, postica valde 

 semicordato-ampliata abrupte angustissiiue decurrentia, medio mar- 

 . gine p istico \-ciliata, vel dentato-2-6-ciliata, caeteruni integerrima, 

 ssepe omuino inermia. Cellulce miuutute, subconfonnes, teq^ui- 

 laterae, subpellucidae, sublfevissitnge, ad angulos iacrassatte, trigo- 

 nis coDspicuis, endochromio parco. 



Foliola adveniunt caulis apicem versus et ad ramos novellos : pr^- 

 minuta, subliaearia, caule multo augustiora, appreusa, ad medium 

 usque inffiqualiter bifida, laciniis subulatis. 



Andrcecia caulis ramive apicem — raro medium — tenentia, spicata ; 

 hractecs, 6-10-jug9e foliis caulinis consecutiva?, perpaulo miaores, 

 turgidae, antice lobulo angusto incubo in folium sensim transeuute, 

 nee in dentem apice protracto, auctse. A.ntheridia baud inveni : 

 sec. el. Carrington 2-3-na sunt. 



Forma foliis omnibus (inlimis ramorum minutis subincisis exceptis) 

 integerrimis legit cl. C. Howie supra Ben Each, Co. Ross. 



The leaves overlap the stem considerably on both sides, a little more 

 on the under than the upper side ; they are very uneij^ual-sided at the 

 base, narrower anteriorly, and running a long way down the stem in 

 the shape of a gradually narrowing wing, but posteriorly much 

 dilated and semicordate, then suddenly contracted into a decurrent 

 base, which is scarcely more than a ridge or slight rim on the stem. 

 In the perfect form (e.g., in Dr. Stirton's specimens from Ben Lawers) 

 there are mostly present about midway of the rounded postical margin 

 a few ciliiform teeth; while the very much smaller leaves towards 

 the base of the stem and branches are always deeply cut, as above 

 described. The cell-walls are decidedly thickened, and have large 

 trigones at the corners ; endochrome sparing ; oil-corpuscules none. 

 In most of these particulars, and in others indicated in the detailed 

 description, there is a close and unmistakable approach to J. colorata 

 and J.grandiflora. But mNardia compressa the leaves project almost 

 equally in front and at back ; they are cordate-reniform, quite entire, 

 almost symmetrical at the base, and not decurrent. The cells are 

 much larger, very slightly thickened at the angles, and they contain, 

 besides the endochrome, a few oil-corpuscules, like those of N. 

 scalaris, but less numerous. 



The branches do not spring from the very middle of the under- 

 side of the stem, but are so far axillary as to be always veiled at the 

 base by the postical angle of the adjacent leaf: i.e., they -dre post ico- 

 lateral. They have the same origin in J. colorata and J. grandifiora. 



. Carrington found the folioles subulate, entire, and so they probably 

 are, occasionally, as in the allied J. colorata. All that I have seen 

 Avere more or less bifid ; the most perfect were composed in the lower 

 half of small quadrate cells, 6 in the length, and 3 (or 4) in the 

 breadth; the laciniae 2 cells wide at the base, at the apex only 1, and 

 o or 4 cells long — these cells being rectangular or oblong. • 



