remains less fragmentary and less difficult to collect and preserve There remain the 

 flowering plants and cryptogams, both of which are extremely useful; of these we have 

 undertaken specially the flowering plants. 



During recent years we have therefore devoted all our available time to the study 

 of the living and newer Tertiary floras, employing methods not in general use for 

 collecting the fossils. We found that the usual method, the study mainly of the fossil 

 deciduous leaves, did not yield althogether satisfactory results, either for the determina* 

 tion of the species, or for the size of the fossil floras. In our research, the evidence being 

 cumulative, we must have a large number of species, and we found it advisable to work 

 chiefly with fruits and seeds, which are also more readily collected, and preserved in a 

 determinable state, than leaves. 



It is evident that a collection of fossil leaves, mainly deciduous, will contain 

 quite a different assemblage from that found in a collection of fossil fruits and seeds; 

 for the latter is not only more varied, but contains also the herbaceous genera, usually 

 absent in palaeobotanical collections. If we desire to study the changes in the flora during 

 successive periods, we must work as far as possible with fossils deposited under similar 

 conditions and collected by the same methods. It is almost impossible yet to make a 

 satisfactory comparison between the flora of a deposit yielding only leaves, and one 

 containing seeds. 



Systematic washing of the seed*bearing deposits in Britain has yielded most 

 interesting results. We have now a good sequence of small floras, commencing with 

 Roman deposits, and going back through Celtic, Neolithic, Glacial, Interglacial, and 

 Early Glacial strata to the latest Pliocene stage, shown in the Cromer Forest*bed. But 

 at this point there is in Britain a break in the succession of land and freshwater strata, 

 the earlier Pliocene deposits being marine and containing no plants. When we again 

 find plankbeds it is so far back, in the earliest Miocene or latest Oligocene, that com* 

 parison is impossible. 



Thus we were forced to look abroad for the continuation of the history of the 

 Pliocene flora of north-western Europe, and when we heard of the discovery of a 

 Pliocene flora in Limburg we immediately arranged to examine it. The study, about 

 seven years ago, of the Upper Pliocene flora of Tegelen yielded results so unexpected 

 as to necessitate a reconsideration of the whole question of the origin of the Palaearctic 

 flora. When therefore Dr. W. J. Jongmans asked us to undertake the examination of 

 material obtained from deposits probably of somewhat earlier date, we willingly con* 

 sented. Perhaps we should not have consented so willingly had we known beforehand 

 how many months of work would be required, or what an enormous mass of material 

 would be sent to us. 



The work of collecting the fossiliferous material in Limburg and in the clay*pits 

 just over the German border was done by Dr. Jongmans and Dr. Tesch. In the Her= 

 barium at Leiden the various gatherings of lignitic material were boiled down with 

 soda, so as to remove the inorganic sediment, and also the decayed humus. Some of the 



