486 BEV. GtEOBGl IIENSLOW ON A THEORETICAL 



connections between Caryophyllece and the " cyelospermous 

 orders ; similarly the Euphorbia cece are considered to be " close 

 with Malvacece" (Hooker), &c. 



On the other hand, while the different orders of endogens 

 have affinities among themselves, no systematist at the present 

 day, I believe, recognizes any real points of connection between 

 endogens and exogens*. Hence arises a difficulty in endeavour- 

 ing to refer the former class to its origin f ; especially as Geology 

 lends us little or no aid in any endeavour to trace the history of 

 its descent %- 



Evidences from Geology. — With reference to any evidence 

 from Geology, it would seem that plants which were at one 

 time supposed to be endogens in the Carboniferous epoch, such 

 as the " palm " JPoaciies cocoina, Lindl. & Hutt., and the 

 '* aroid " Eothocites, really belonged to gymnosperms, and that 



true endogens have not been found earlier than the Trias or 

 Permian §. 



M. 8a porta has described || some fossil plants found in strata 

 above the later Jurassic series at Cereal in Portugal. They 

 appear to have been aquatic in habit. One of the more remark- 

 able is called Protolemna, and resembled existing Lemnce in 



* Robert Brown considered Cycadete to be in certain respects " transituin a 

 Monocotyledonibus ad Dicotyledones," and places it in Monocotyledons next 

 to Hydroeharidece. He also thought Aristolochia to be " affinitas arctior cum 

 monocotyledonibus quibusdam " ; although, following Jussieu, he places it in 

 the Dicotyledons (Prod. Fl. Nov- Holl. pp. 347, 349). 



t Prof. Delpino remarks: — "Le Aere origini delle monocotiledoni sono 

 tuttora avvolte in quasi completa oscurita" (Smilacese, p. 51, note). 



% -Although I had long ago come to the general conclusion, but had not 

 collected sufficient materials, that endogens were in some way connected with 

 \%ater, I was glad to see the following observation by Mr. W. Gardner in his 

 paper on Watei-ylands (Proe. Camb. Phil. Soc. 1883, p. 43):— "I should like 

 to put forward a view which has struck me, but upon which I would prefer to 

 lay no stress. Comparing generally the Dicotyledons with the Monocotyledons, 

 it seems to me that whereas the former are typically land-plants in their habit, 

 the latter on the other hand are of essentially an aquatic nature." 



§ Thus one of the earliest plants, described by Schimper, occurs in the Trias, 

 and is called JEtivphyllum. It cannot be assigned to any existing order. It 

 seems to have had a branching woody stem, but one more like that of a grass, 

 cgreeing so far rather with bamboos than anything else, and bearing spikelets 

 which were articulated, as is the case with many grasses now. It, however, 

 appears to have been devoid of nr;des ; in which feature it resembled Typka. 

 (• Trait* de Paleontologie V^getale,' vol. ii. p. 386.) 



1 Comptes Rend us, cxiii. 1891, p. 249. 



