iS96. SOME NEW BOOKS. i35 



we think the following-up of cell lineage has been too exclusively 

 pursued, at the expense of a wider morphological survey of the 

 mature forms and of the numerous adaptations to circumstances of 

 the environment so fully illustrated by these plants. 



Naturally, the vascular cryptogams occupy the largest share of the 

 book. Here Professor Campbell works out the views, which he was 

 the first to propound, and which have been largely accepted by many 

 other botanists, especially in this country, viz., that the eusporangiate 

 ferns are older, phylogenetically speaking, than the leptosporangiate 

 forms. The author also insists that it is among the Hepaticae, 

 rather than the Musci, that their lower affinities are most naturally 

 to be discerned. 



The Ophioglossaceae are regarded as the most primitive type, and 

 as being most nearly related to such a form as indicated by Anthoceros. 

 The author works out this point, so far as the evidence goes, both in 

 the sporophyte and the gametophyte, and he lays considerable weight 

 on the results of comparisons between the antheridia and archegonia 

 of the two groups respectively. We confess that, as regards the 

 antheridia, we do not think that a very good case has been made 

 out, nor indeed does Professor Campbell express himself on this 

 point without reserve. The antheridia of the Anthoceroteae are very 

 peculiar, and need some manipulation in order to be forced into line 

 with those of the ferns. However, it must, of course, be remembered 

 that, in any case, it is not Anthoceros itself which is contemplated as 

 lying in the line of direct descent ; it only serves to give the idea of 

 a generalised type to which the ancestors of the present higher plants 

 may have roughly conformed. The argument drawn from the com- 

 parative study of the archegonium is much more convincing, and the 

 author's suggestion that the whole archegonium of a vascular crypto- 

 gam corresponds only to the axile series of the analogous body in 

 liverworts we regard as a peculiarly happy one. 



The leptosporangiate ferns, so far, at least, as concerns the 

 homosporous members, form a well-defined group, their relationship 

 with the eusporangiate series being indicated through the Osmundaceae. 

 As regards the heterosporous families, Professor Campbell con- 

 siders the Salviniaceae as a possible offshoot from the already 

 divergent Hymenophyllaceae, while the Marsiliaceae are assumed to 

 be derived from the Polypodiaceae. We should ourselves, however,^ 

 be disposed to remove the Marsiliaceae much further from any 

 existing homosporous type than Campbell has done; the very strongly- 

 marked degree of reduction in their gametophyte and the great 

 peculiarity of the sporophyte, botbin structure and in position, seem 

 to us to be facts which indicate that, if they are really related to the 

 Polypodiaceae at all, their affinity is at best but an extremely remote 

 one. It is, we venture to think, not impossible that they may turn 

 out, after all, to be more nearly allied with some eusporangiate 

 prototype resembling the Ophioglossaceae. It is not a priori im- 

 possible that the present character of their sporangia may well have 

 been developed quite independently of any real descent from 

 primitive leptosporangiate ancestors. Heterospory has appeared 

 more than once in the alliance of vascular cryptogams, why not the 

 leptosporangiate character also ? 



The general summary at the end of the book is a fine piece of 

 work. The full discussion of the mutual relationships existing between 

 the various subdivisions of the cryptogamic archegoniatae, and between 

 these and the higher spermophytes, is full of interest and suggestive 

 thought from first to last. Professor Campbell regards the phanero- 



