56 DESCRIPTIVE CATALOGUE 



pollen-grains, form separate and distinct cones. Male cells non- 

 ciliated, and almost passive within the pollen-tube. Integu- 

 mented seeds borne unenclosed upon open scales, which, however, 

 may overlap so as to hide thorn. Embryos either dicotyledonous 

 or polycotyledonous. 



Casts or imperfectly petrified $ cones, casts or impressions 

 of leaves, and decayed or well-petrified portions of the secondary 

 wood of main trunks and brandies, form the principal fossil 

 remains of this group, of which a good general account is given 

 by Seward (1895, pp. 185 et seq.X 



In the Lower Greensand deposits at present being described, 

 foliage twigs and cones are both exceedingly rare and can easily 

 be dealt with under their respective families ; fossil woods, on 

 the other hand, are present in large numbers and offer a complex 

 problem. 



The researches touching the determination of fossil coniferous 

 woods date from the beginning of last century, and are very 

 extensive. A complete survey of the literature cannot be 

 attempted here, but the following notes indicate some of the 

 more important papers. 



Hartig (1848) is given the first place in the historical sequence 

 by Kraus (18(54), who remarks on the accuracy of much of his 

 detailed observation. Goeppeft'a famous monograph, which is 

 the basis for several of the leading genera, was published in 

 1850. Outstanding, not only among the earlier writers, but 

 among all work on the subject, is Mercklin'a (1855) monograph 

 for his magnificent illustrations of the exact appearance of the 

 minutest details of a number of woods. In many of his drawings 

 he illustrates well the pitting of the medullary ray-cells, a 

 feature often overlooked by more recent workers. 



Kraus made important investigations in 1864, 1865, and 1866, 

 and in a section of Schimper's ' Paleontologie ' (1870) established 

 five of the main groups in which fossil coniferous woods are still 

 placed, though his system has been considerably extended and 

 further subdivided by Schroter (1880), Gothan (1905, etc.), and 

 others. 



Among papers of less significance about this time and even 

 later, much stress was often laid on the width of the annual 

 rings a very deceptive character. As Con went/ ( !S7f, p. lM) 



