694 THE AMERICAX XA TL'RAIJST. [\ ( .l. XXXVI II. 



arise in the roots of Ahicx balsauiea which have heen injured, 



confirming the observations of Anderson that tliey may be trau- 

 matic in their origin. The most significant facts, however, 

 relate to the normal occurrence of such cysts in Sequoia. He 

 shows in the first instance that they are absent from the wood 

 of the first year's growth in Sequoia senipcrviirus, while they 

 are present for the same period of growth in .V. gigantca, though 

 absent from the growth of later years. In both species they 

 arise in the earlier spring wood. Jeffrey concludes that the 

 tangentially disposed resin cysts of Sequoia sempervirens repre- 

 sent the result of injury, and he would apply this rule to all 

 similar cases in the various species of Abies and Tsuga. 



Some years since, De Bary (9, pp. 490 and 495) formulated 

 the law that " resin canals .... occur in the ligneous bundles of 

 the same Abietineae which possess horizontal canals in the med- 

 ullary rays." This is a law^ of very great constancy, and, as 

 recently shown by Penhallow (39), it is applicable without 

 exception to all living species. But as the same author (41, 

 p. 42) has shown more recently. Sequoia burgcssii from the 

 Eocene of the Northwest Territories offers a remarkable excep- 

 tion to this law, since it shows well defined resin passages in the 

 medullary rays, but without corresponding structures in the vas- 

 cular bundles. 



As presented by existing species, Pseudotsuga, Larix, Picea 

 and Pinus, without exception, show resin passages in both the 

 radial and longitudinal positions. In transverse section they are 

 scattered throughout, sometimes appearing chiefly in the sum- 

 mer wood, sometimes chiefly in the spring wood, or again about 

 equally in the two regions, and they rarely conform to the pre- 

 cise law stated by De Bary (9, p. 495), that - they lie scattering 

 in a ring in the external region of every annular layer." The 

 constancy of their occurrence in the four genera mentioned m- 

 volves very few features which call for special comment. In 

 Pseudotsuga and Larix the resin pas.sages are scattering. 

 Sometimes they not infrequently unite in pairs so as to form 

 short, tangential series and they thus approach the type of 

 Tsuga or Abies, while yet again they may become definitely iso- 



