286 



NATURE 



\Aug. 3, 1876 



SMITH ON FERNS 

 Historia Filiciimj an Exposition of the Nature., Number, 

 and Organography of Ferns. By Jno. Smith, A.L S., 

 Ex-Curator of the Royal Botanic Garden, Kevv. With 

 Thirty Lithographic Plates by Fitch. 8\ro, 429 pp. 

 (London : Macmillan and Co.) 



THE main and most valuable part of this work is a 

 full account of Mr. Smith's scheme of fern-classifica- 

 tion, with a complete catalogue of all the known species, 

 arranged according to his views and diagnostic characters 

 of all groups of a higher grade. The author is the 

 patriarch of living fern writers, having worked at ferns 

 with unwearied perseverance and enthusiasm for now 

 upwards of fifty years. In 1823, when he first took charge 

 of the living collection at Kew, it contained only forty 

 species. Sir Wm. Hooker also, as is well known, made 

 ferns his favourite department of botany for the last 

 twenty-five years of his life. In 1846 the living collec- 

 tion had increased to 400, and in 1857 to 600 species. In 

 1864, when in consequence of failing eyesight Mr. Smith 

 was compelled to resign his appointment, he estimated 

 the number of ferns in cultivation in the country at up- 

 wards of 1,000. The whole number of species now known 

 in the world, taking a broad view of what constitutes a 

 species, is not far short of 3,000, and during the last year 

 certainly not less than fifty new ones have been added to 

 the list. 



The great peculiarity in Mr. Smith's plan of fern-classifi- 

 cation is that at the outset he divides ferns into two groups, 

 which he calls Desmobrya and Eremobrya, an account of 

 which will be found at p. 65. The difference between them 

 depends mainly upon whether the stipes are continuous 

 with the caudex, or jointed at the base, so that they 

 become detached when the frond withers, like the leaf of 

 one of our deciduous trees. The Eremobrya, which are 

 comparatively few in number, are such ferns as Folypodi- 

 um vulgare, and Davallia canariensis, in which the fronds 

 are produced singly from the sides of a creeping rhizome 

 and are jointed at the base. The Desmobrya, which are 

 perhaps three quarters of the family, and have unjointed 

 stems, may have either the fronds produced in a crown 

 from the summit of an erect caudex, as in the tree-ferns 

 and Nephrodium Filix-mas, or produced alternately in a 

 single series from a creeping rhizome, as in Pteris aqui- 

 lina and Nephrodium Thely pteris. These last, which are 

 comparatively few in number, are like the Eremobrya in 

 habit, but want the joint. 



The old Swartzian and Willdenerian genera, founded 

 upon the shape and position of the sori, and the absence 

 or presence and position of the indusium, fall many of 

 them partly into one of these groups, partly into the other, 

 and this holds good also with ferns in which sori and 

 veining also coincide. So that there are substantially 

 three plans of fern-classification and fern-nomenclature, 

 each of which is represented by a recent work in this 

 country, and their relation to one another is as follows : — 

 All systematists agree in recognising a substantial differ- 

 ence in the shape and structure of the sporangia, the 

 shape and position of the sori, and the absence or presence 

 of an indusium as constituting a genus. In Hooker and 

 Baker's " Synopsis Filicum," now in its second edition, 

 only genera are admitted which rest on these characters, 

 and their number is 76, Polypodium, containing about 



400, and Asplenium about 300 species. There is great 

 variation in the arrangement of the vascular bundles in 

 the fronds of ferns. Sometimes they do not join again 

 after once branching. In other cases they join and form 

 meshes of various shapes. A second school, represented 

 in Britain by Moore's " Index Filicum," regard any 

 app^reciable difference in veining as constituting a generic 

 character, and this increases the number of genera 

 between two and threefold. The total number of genera 

 admitted by Moore is 178, and of these, twelve go into the 

 Polypodium of Hooker. Mr. Smith's plan carries us a 

 decided step further in the direction of subdivision, and 

 by using the character already explained as a ground of 

 generic separation, raises the number of genera admitted 

 to 220. But in point of fact all the ferns in which 

 the sporange is surrounded by an incomplete ver- 

 tical ring (Polypodiaceae), v/hich are three-quarters of 

 the whole order, agree completely in the essential struc- 

 ture of their organs of nutrition and reproduction, so 

 that a large proportion of the genera even of those 

 that admit the fewest number are separated from one 

 another by very unimportant characters, and the great 

 difference that there is in the nomenclature of ferns 

 according to the three different systems does not repre- 

 sent any deep-seated divergence of view, because the 

 systematists of the first school willingly accept the further 

 subdivisions of those that multiply the number of genera, 

 as being the best possible groups that can be devised of 

 subgeneric or sectional value. The book, therefore, is 

 worthy of careful study by everyone who is interested in 

 the subject ; it is a complete gathering together in one 

 view of the author's work in the field to which it relates. 

 Remembering how the book has been written, no one can 

 study it without strongly sympathising with the author in 

 the difficulties under which he has rested in thus placing 

 before the world the matured result of his labours, and 

 admiring the energy with which he has achieved so diffi- 

 cult a task in such trying circumstances. 



In the way of criticism we have two observations to 

 make : the first, that whoever has undertaken the correct- 

 ing for the press has done his work the reverse of well. 

 Names of well-known genera, species, authors, and 

 books are frequently misspelt. At p. 65 we have the 

 essential character of Desmobrya made to depend upon 

 venation, and at pp. 98 to loi we have under both Nipho- 

 bolus and Colysis all the three genders represented in 

 the adjectival specific names. Secondly Mr-. Smith, fre- 

 quently under a genus, compares the number of species 

 as given in Hooker's " Species Filicum " with that given 

 in Hooker and Baker's " Synopsis Filicum," as if the 

 two numbers represented the same thing. Under Adian- 

 tum, for instance, he expressly says that where Sir W. 

 Hooker has made 109 species Mr. Baker has reduced 

 them to sixty-two. He has evidently forgotten that, as 

 was fully explained in the preface to the later work, the 

 plan of the two books is different — that the more con- 

 densed " Synopsis " only includes the species known with 

 certainty by the authors ; but the " Species," in addition 

 those that have been described by others, but not 

 identified, a large mass of doubtful plants in addition to 

 those that are known fully and clearly, so that the two 

 sets of figures cannot be fairly compared unless this be 

 constantly borne in mind. J. G. B. 



