VOL. III.] Recent Literature. 77 



anterior to the Mesozoic The remains occurring in older forma- 

 tions, and usually attributed to this class, are either too imperfect to 

 permit of certain classification, or may better be referred to other 

 groups. Of living genera Araucaria is certainly known as far back 

 as the Jurassic, and Sequoia as the Cretaceous; Ginkgo is still older. 



A special chapter is devoted to the Cycads and Medullosae, and 

 another to the remarkable entirely extinct group of the Cordaiteae. 

 To the latter, which are separated entirely from the Coniferae, are 

 referred many of the remains of fossil wood which have usually been 

 supposed to belong to the Coniferae. The most interesting point in 

 connection with them is the discovery of flowers, both male and fe- 

 male, in a sufficiently perfect state of preservation to give a very lair 

 idea of their structure, which differed materially from that of any 

 living gymnosperms. The pollen grains are preserved with re- 

 markable perfectness, even showing a group of cells within which is 

 assumed to be a sort of rudimentary prothallium like that in the 

 pollen of other gymnosperms, but much more highly developed. 

 These points seem to warrant the separation of the Cordaiteae as a 

 class, co-ordinate with the Cycads and Conifers. 



The chapter on the ferns is especially interesting and suggestive. 

 While a considerable number of ferns have been found with well 

 preserved fructification, all of these in the formations below the Me- 

 sozoic, that can be positively determined, show affinities with the 

 Marattiace^e and, perhaps, with the Ophioglossese. This fact is es- 

 pecially significant, as it entirely reverses the ordinarily accepted 

 arrangement of the leptosporangiate and eusporangiate ferns. The 

 former — i. e., those ferns in which the sporangia are of strictly epi- 

 dermal origin — are usually regarded as the simpler forms from which 

 the Eusporangiatae, or those forms with massive sporangia, like the 

 Marattiaceae and Ophioglosseae, have been derived. 



As the Leptosporangiatae have firm sporangia that ought to have 

 been preserved in a fossil state, it is difficult to account for their ab- 

 sence from the coal measures and earlier formations, if they really 

 existed when these were forming. It seems probable that they are 

 really later, more specialized forms, derived secondarily from the 

 more primitive Eusporangiatae. This view accords, too, with the 

 evidences of embryology, and simplifies very much the problem of 

 the origin of the phanerogams. The remains of hydropterides are 

 very scanty, and only a few remains from the tertiary are beyond 

 dispute. 



