Structure of Recent and Fossil Conifer a. 313 



having sustained slight and very irregular interruptions. On 

 cutting a thin transverse section, the slender cohesion of the 

 radial partitions is equally striking as in the Moreton Bay 

 Araucaria, it being equally difficult to cut one, even of small 

 extent, without these partitions separating from each other. 

 The meshes of the transverse section are also very irregular in 

 respect of form. But whilst in these particulars the two kinds 

 of wood agree, they differ most widely, as seen in the longitu- 

 dinal section parallel to a radius. In the CalHtris, the discs 

 which occur both in single and double rows, resemble those oc 

 curring in many of the true pines. In the double rows they 

 are placed side by side as in ihe pines, and whether in single 

 or double rows they are similar in size as well as in form. 



The Pinus lanceolata (Cunninghamia lanceolata of Brown) 

 so greatly resembles, in its general aspect, as well as in the form 

 of its leaves, the tribe of Araucaria, that I was led to expect that 

 its internal structure would be somewhat similar. On cutting 

 the requisite sections, however, a very different result was ob- 

 tained. The transverse section presents well defined annual 

 layers. The quadrangular form of the meshes predominates, 

 and, as in many of the true pines, the meshes gradually dimi- 

 nish in size towards the outer limit of each annual layer. The 

 longitudinal section parallel to a radius, has discs similar to those 

 of the pines, and the concentric longitudinal section has expan- 

 sions in the partitions equally similar. 



In my former communication on this subject I stated, and 

 that, too, entirely from the anatomical structure of the wood, that 

 the Salisburia adiantifoUa belonged to the coniferae. I have 

 since been informed that botanists at present maintain that it 

 does not. At a former period it was classed with the coniferae, 

 and it is not impossible that, at some future period, some one 

 may again restore it to its former place. At all events there can 

 be no dispute as to the structure of its wood being similar to 

 that of the trees which botanists have styled coniferae; and that 

 those who have not had an opportunity of examining the wood 

 may be enabled to form some ideas of its structure, there is in- 

 troduced into Plate V. representations of the three usual sections. 

 Figure 3 is transverse, 4 is longitudinal parallel to a radius, and ^ 

 5 is longitudinal concentric. 



