548 THE AMERICAX XATL'RALIST. [\'ol. XXXVIII. 



the first three genera and the last, in such a way that the former 

 represent an intermediate, while the latter represents a completed 

 type, it will be necessary to examine them separately. In all of 

 these cases, however, the secretory reservoir is invariably char- 

 acterized by the presence of a definite and continuous canal 

 of indeterminate length, in consequence of which I reserve for 

 them the appropriate and long used term, "resin passage," as 

 distinguished from the resin cysts of the previously discussed 

 genera. 



In Pseudotsuga the resin passages are always scattering, 

 though they frequently occur in tangentially extended groups of 

 two or four contiguous or even coalescent reservoirs. The 

 central canal, which is usually small and not infrequently very 

 narrow, is rather more generally rounded than in previous types. 

 The epithelium is very clearly defined and consists of one to 

 three rows of thick-walled parenchyma cells, sometimes con- 

 taining resin, the first row of which are rather small and radially 

 flattened, but in P. macrocm-pa they are rather thin-walled. In 

 P. donglasii the epithelium is commonl}^ extended on the two 

 sides of the resin canal in such a way as to form a tangentially 

 elongated tract which not infrequently extends beyond and 



other hand, the epithelium is concentric with the canal, thus 

 forming a tract of about equal thickness all around. Such a 

 deviation as is expressed in P. donglasii constitutes the first 

 c\ i(lcncc of a tendency in development which is fully and fre- 

 quently expressed in Pinus. Thyloses are of infrequent occur- 

 rence, and they appear to be confined to P. viacrocarpa, where 

 they are few in number and generally rather thin-walled. Paren- 

 chyma tracheids are usually not apparent in a transverse sec- 

 tion. This results from the frequent location of the resin 

 passages in the summer wood, which is not favorable to their 

 dexclopmcnt, and from the close resemblance which they bear 

 to the tracheids of the spring wood; and while such elements 

 torm an integral part of the resin passage, their particular dis- 

 I '-sit ion cannot be exactly defined, though there is no good 

 i^ason tor supposing that they differ in this respect from what 

 may be observed in other cases. In a longitudinal section the 



