276 30 



between /). oligocarpa and its nearest allie« D. pilosnla, D. argentina and D. Nockinnn 

 see undei' these species. 



The Brazilian forms which I now refer to D. oligocarpa are much more 

 varying. I have had much correspondence with Dr. Rosenstock on the right con- 

 ception of these forms, and this keen pteridologist is of opinion that in Southern 

 Brazil there occur as least two good species, of which one may be D. oligocarpa. 

 His conception may be and is probably right, but examining again and again the 

 numercnis si)ecimens jilaced before me, I find it im])ossible to draw a line between 

 the forms themselves and between them and the true D. oligocarpa. I shall there- 

 fore confine myself to giving some remarks on the different forms. 



Three names are applied to these forms from Southern Brazil: 



1. Polgpodiam retiisiim Sw. Vet. Akad. Handl. 1817: 61; Lindman, Arkiv l'or Bot. 

 1: 227 tab. 10 fig. 11; Dryopteris retusa C. Chr. hidex 288. 1905. 



Type in HS leg. Freyreis! 



2. Pohjpodium piibescens Raddi, PI. Bras. 1: 23, tab. 34. 1825 (non L.). 



3. Aspidiuin Kaulfussii Link, Hort. Berol. 2: 117. 1833('.'); Mett. Fil. Lips. 90. 1856; 

 'f Nephrodium Kaulfussii HK. sp. 4: 97. 1862; ?Bak. Syn. 268; Dryopteris Kaulfussii 

 O.Ktze. Rev. Gen. pi. 2: 813. 1891. 



This last, founded on a cultivated specimen, belongs no doubt here, at least 

 it is the case with the specimen described by Mettenius under this name (HB!), 

 while it may be somewhat doubtful whether Link's plant is the same. Nephrodium 

 Kaulfussii Ilk. & Bak. is from the description quite undeterminable. The character 

 given in this: sori medial, does not agree witli any form of the whole series. 



Having no authentic specimen of Pol. pubescens Raddi it is, of course, impos- 

 sible to say exactly, which form this name was applied to, but judging from Radoi's 

 rather good figure his plant no doubt belongs here and is very nearly allied to 

 Pol. retusum Sw., if not the same. Dr. Rosenstock points out the following dille- 

 rences between the two forms: "Die Raddi'sche Pflanze ist etwas länger gestielt, 

 ihre Lamina ist nach unten breiter als P. retusum, und die Segmente 2.0. sind 

 an der Spitze nicht gestutzt, stehen auch etwas entfernter von einander als bei 

 retusum" (Rosenstock in litt.). On the other hand the character attributed by 

 Rosenstock (Hedwigia 46: 122. 1906) to his U. retusa (Sw.) uar. austrohrasilieusis, 

 which he believes identical with D. puhcscciis Raddi: "facie frondis inferiore pilis 

 bamatis undique obsita a typo diversa", cannot l)e upheld, as the original P. retu- 

 sum shows the same character to a less degree. But there is a conspicuous tlifie- 

 rence in the shape of the segments, which in retusum are obtuse (see Lindman's 

 figure) in pubescens mvich more acute. Still it seems to me that this difference is 

 only of minor importance, as I find a great variation in the specimens examined 

 as to that character. With regard to the presence of glands on the underside of 

 the lamina I find also a similar variation; while some specimens are very finely 

 giandulose, as in P. retusum, others are densely glandidose and others again entirely 

 eglandulose. Also the reduction of the leaf towards its base and pubescence of 



