524 



\MERI 



I TURALISr. [Vol. XXXVI 1 



our investigations, are wholly devoid of such structures, in the 

 sense defined above. 



In Arancaria excels a a transverse section shows more or less 

 numerous elements containing resin. These are not to be dis- 

 tinguished in their general structure from the surrounding tra- 

 cheids, and they are to be recognized solely by their contents, 

 which are usually somewhat prominent. Their distribution is 

 characteristic. They occur in small, scattered groups, or more 

 commonly in rows one or two elements wide, parallel with the 

 medullary rays and in immediate contact with them on each side. 

 When the plane of section passes near the position of the sup- 

 posed terminal walls the latter may be seen to be cut through in 

 various ways, but they never exhibit any structural features, and 

 they are therefore in no way compar- 

 able with the terminal walls of the 

 wood-parenchyma cells. In a radial 

 section they are seen to be long and 

 fusiform, exactly resembling the wood 

 tracheids, except for a reddish brown 

 transverse plate which occurs either 

 close to or exactly opposite a medul- 

 lary ray, a position .which is more 

 clearly shown in a tangential section 

 (Fig. 35)- The dark plates closely 

 resemble Sanio's bands, for which 

 they might very readily be mistaken 

 upon casual observation, or they might 

 likewise be mistaken for terminal and 

 unpitted walls. In Agathis aiistralis 

 these features are represented in their 

 typical form. The transverse section 

 Fig. ^i.-Agathu austraiis. Trans- shows such elcmcnts to bc numcrous 

 of the TesTnousTrachdds ^Ipp"" disposed" iu radial rows on each 



site sides^ of the medullary ray at r. side of the medullary ray (Fig. 32). 



In a radial section they present the 

 same fibrous and fusiform character as in Araucaria, but in 

 addition the wall usually experiences a marked increase in 

 secondary growth within a region exactly opposite the ray (Fig. 



