[1'KNitAi.i.ou] NORTH A.MKKICAX Sl'KCIKS OF DADoXYLON 39 



Tlio i)0fuliiir distribution of tlio bordered piks constitutes one of 

 the most prominent ami distin-uisliin- features of the genus, and dis- 

 tinctly alhes it to tlie .VI•aueclriea^ L'nger hiys down iis a genenc char- 

 a^-ter, "Vasa porosa, poris in series 1— i spiraliter dispositis,"' which is 

 simply a restatement of JJrongniarfs recognition that "les p..netuations 

 des libres hgneuses disposees en plu^sieurs series alternautes entre elle, 

 et prenant par pression la forme d'areoles hexagonales." All studies 

 show that the ],ils are distributed througlu.ut the length of the 

 tracheids, a< well as that they regularly occur on the radiafwalls of all 

 tracheids, thus presenting an iniportimt contrast to the law of distril.u- 

 tu)n which excludes them more or less fully from the tracheids of the 

 suntmer wood of the AbietiiKw. 



\Vhile these laws of distribution are well recognised as ap|.lic;ible 

 in a generic sense, the various recognised species show great diversity 

 111 this respect, und such variations have l>een taken as an element in 

 diagnosis. The possible value of such data is fully exemplified in the 

 case of C. Xewberryi, where the pits occttr in well defined groups which 

 fall into radial rows. This feature ha.s ttnforttmately been neglected 

 entirely, by every writer exc<>pt Prof. Xewberry himself, but its im- 

 port'ance as a, diiferential character cannot be doubted (tig. U). It is 

 nevertheless a well recognised fact that the number of rows in wb.icli 

 pits occur is such an altogether variable factor within the same in- 

 dividual, as to render it untrustworthy as a differential character except 

 when employed with great caution.^' Under these circumstances we can 

 only consider such data as of senice in so far as they help to reinforce 

 other eh'aracters indicative of specific differences, while at the same time 

 they tend to enforce the idea that one should be very cautious in 

 creating new species unless such can be justified bv well defined 

 characters. 



The medullary rays refpiire to be disctissed fro.n two points of 

 ^-iew. The tangential exposure shows the cells to be disposed in one 

 row, becoming ;.^-soriate in ]iart, and Kraus h^is employed this feature 

 as a means of diiferentiation from those other genera in which multi- 

 seriate rays occur.'' In a radial section the rays are scnm to be C(nnp(">sed 

 of one kind of element oidy. with the exception of C. Clarkii, in which 

 definite tracheids are to be found in addition to the ordinary 'parenchy- 

 matous elements. This is the ""l^^speci es within my o bservation, which 

 ' Gen. et Spec. PI. Foss., 1850, 378. " ~ ~~ 



V i"c?; T!^ ^°'''' ^°*- ^^'^•' ^^*'^' 22^= ^^"«' Wurzburger Naturwiss. Zeit., 

 v., 1864, 385. 



' Schimper, Pal. Veg., 379. 



