No. 475] REVISION OF COMPTONIA 



497 



Dawson's more or less, are Conipfonlphi/Uum japonicum Xath., 

 and especiary Comptonia parfita L(\s(|., tlie latter l)ase(l i]j)()ii a 

 poorly drawn fragment from the (Ireen Uiver beds. 



The existing Comptonia sometiines fiiniishes leaves very similar 

 to diformr, atid 1 have c.^lh-cted many such, althougli usually they 

 differ in l>eiiig somewhat hroaih'r, r. ;/., PI. L>, Fig. 4. 



The many exceIl(Mit figures |)ul>lishe(l l)y Kttingshausen furnish 

 adequate and ty{)ical examples of the leaves of tiie s|)ecies imder 

 di.scussion. This form of leaf makes its appearanee during the 

 Eocene at the widely separated localities of the Isle of Wight in 

 England, Brives in France (represented by C. vinai/i Sap.), and 

 Murderer's Creek in New Zealand. It is possible that the 

 New Zealand leaf, which is identical in form, may have been 

 borne on an entirely different plant, as it is difficult to accotim 

 for so wide a distribution. It may well be that the New 

 Zealand leaf should be referred to Dryandra as Ettingshauscii lia^ 

 done, since several species of the latter genus hnw been reconieil 

 from that region by the same author, all having leaves of the aruli- 

 loha type. In discussions of the distril)uti(>n of Com{)t()nia we 

 should constantly bear in mind, however, tlu- ineagernos of the 

 record of the upper Cfetaceons period. 



Comptonia gaudinii I leer 



