312 BOTANICAL GAZETTE [October 



3. Simplification of a compound sporophyll has been attained 

 to fullest extent in Arthrotaxis selaginoides, Agathis, and Saxego- 

 thaea, and possibly others; an extensive reduction of bract occurs 

 in Cedrus Libani and the lower sporophylls of Pinus maritima; the 

 scale in Phyllocladus is probably reduced so as to be represented 

 only by a distinct ovular supply; the welding of the two organs is' 

 complete in Juniperus communis and Chamaecyparis Lawsoniana. 



4. Fusion of bract and scale vascular supplies does not directly 

 parallel fusion of bract and scale. 



5. Separate origin of bract and scale vascular supplies occurs 

 most generally in the Podocarpineae and Abietineae; fusion of 

 bract and scale supplies has reached its highest expression in the 

 Araucarineae; both types of bundle origin are represented in the 

 same strobilus in Cryplomcria japonica, Cupressus Bcnthamii, and 

 the lower sporophylls of Pinus. 



6. The bract bundle in plants with uninerved vegetative leaves 

 divides only slightly if at all; the extent of the scale bundle system 

 is directly related to the size of the organ supplied. 



7. The scale bundles in the Abietineae and Chamaecyparis 

 Lawsoniana form in the expanded portion of the organ a straight 

 row or arc; in members of the Taxodineae and Cupressineae scale 

 bundles swing around so as to lie at each side of the bract bundle. 



8. In Cryplomcria japonica and Cupressus Benthamii and per- 

 haps Cunninghamia Davidiana scale bundles accompany the bract 

 bundle into the free portion of the bract. 



9. A branching bundle in the vegetative leaf in Araucaria and 

 Agathis probably implies a branching bundle in the bract of the 

 sporophyll; the vascular system in the megasporophyll is probably 

 a complex of bract and scale bundles. 



10. In species of Podocarpns the scale bundles continue in the 

 portion of the scale folded toward the dorsal side, forming the 

 epimatium of the ovule. 



The author is indebted to Professor J. M. Coulter and Dr. 

 C. J. Chamberlain for valuable suggestions, and to the latter also 

 for excellent material. 



University of Chicago 



