99 151 
rent, but I cannot consider it specifically distinct. SrÜBEL's nr. 751 from Ecuador 
shows to some extent the same peculiarities and even in the authentical specimen 
of the type one can find a difference of some few millimeters between the length 
of the segments of the upper and lower side of the lower pinnz, 
162. D. Glaziovii Christ; C. Chr. Revision nr. 62 fig. 40, 
Area: Brazil, Minas Geraes and Rio. 
Since the publication of my Revision I have seen the type-specimen of this species, 
Graziou nr. 5267 (C) and specimens from Minas Geraes, Itacolumi, 1400 m, 
ScHWACKE nr. 14109 (C). In these specimens the veins are simple and the basa! 
segments not free; the stipe bears 3—4 pairs of glanduliform pinnz. Our specimen 
(H) has furcate veins and the lower segments often quite free, thus belonging to a 
more divided form that like the preceding species has the basiscop side of the 
pinne enlarged. But the species can be much more divided, fully bipinnate or 
even tripinnatifid. The bipinnate form with simple or furcate veins is 
Gymnogramme patula Fée, Cr. vasc. Brés. 1: 59 tab. 14 fig. 3 1869. 
Serra os Orgaos, Graziou nr. 2822 (H); Serra de Ouro Preto, SCHWACKE nr. 
15022 (C). FÉrE's figure is as a whole excellent, but the lower pinnulz should be 
reduced gradually. This form is closely allied to the Andine D. pteroidea to which 
species BakEn in Flor. bras. referred it; it differs by the sori being placed not so 
close to the margin and not being confluent. Unfortunately I have not seen the 
base of the lamina, but if I am right in considering it a form of D. Glaziovii it 
very likely has the lower pinnz abortive and glanduliform. 
A still more cut form has the pinnule deeply lobed and the veins pinnatifid 
in the lobes; this is 
Gymnogramme expansa Fée, Cr. vasc. Brés, 1: 60 tab. 14 fig. 4. 1869. 
Another specimen of Grazriov's nr. 2822 (H) agrees perfectly with FÉE's figure; 
it is sterile. 
Comparing Fée's two figures with that of mine we see here a series of forms 
quite corresponding to those of D. multiformis. 
While the bipinnatifid forms of the last two species, D. euchlora and D. Glazovii 
certainly are related to D. piloso-hispida the species are, on the other hand, also 
and perhaps more intimately related to the bipinnate D. pleroidea. Together with 
this they belong to a separate little group of large species of rather thin texture, 
and dark colour. Most species belonging here have a nearly entirely glabrous 
frond, opposite and sessile pinne, which are narrowed towards their base, where a 
distinct aérophore is to be found. D. euchlora stands next to D. piloso-hispida and 
resembles not a little the bipinnatifid form of D. Glaziovii while the more cut 
forms of this latter species are closely allied to the two following species, which 
are, as it seems, constantly bipinnate. They represent the highest development of 
culting in the subgenus, at least as for as American species are concerned. 
20* 
