BOTANICAL GAZETTE. 



241 



6. The thorns of Robinia are known to be stipules ; those of 

 Gleditschia, branches; those of some Astragali, petioles; those 

 of the Barberry the veins of the leaves. (See Gray Struct. Bot.) 



7. A study of the Osage orange, Madura aurantiaca, shows 

 its thorns to be transformed branches. Tiny notches toward the 

 tip often indicate nodes on this branch. At the base, on either 

 side, are two scales, only one of which subtends a bud, the other 

 being empty. These are at the two lower nodes of the metamor- 

 phosed branch. On growth of the'young bud into a branch it 

 pushes the thorn aside, so that this appears to be axillary to the 

 young branch in the axis of the really empty scale at its side. 



8. The spines of Ribes Cynosbati are mere outgrowths of the 

 bark, as are also those of the Smilax above mentioned, and there- 

 fore are of no morphological importance. 



9. In Sambueus Canadensis the buds consist of short but 

 broad scales which already subtend several axillary buds, as well 

 as aid in protecting the large central bud which is to furnish the 

 main shoot. A single bud, however, which sometimes appears 

 beneath the central bud, seems to have no connection with it, 

 and I can not find that it is subtended by a scale. It however 

 originates in the same tissues with the larger bud, and the con- 

 nection with it may be traced beneath the bark. Can they, in 

 such cases, be strictly said to be superposed ? 



10. In Lindera Benzoin two branches, superposed in the axils 

 of last year's leaves, each terminate in a leaf bud and have two 

 lateral branches which are flower buds. In spring the lateral 

 buds blossom, and the terminal buds are at liberty to develop 

 into branches, but the upper one generally gets the start, and the 

 other dies. In this way we have extra-axillary branches. 



11. The development of the bud scales of the dogwoods in 

 spring is quite interesting. In Cornus fiorida each scale repre- 

 sents the tip of a leaf, and in spring it shows its character by de- 



• veloping at the base into a broad petaloid blade, which bears 

 the withered scale at its tip. In this condition the scales serve 

 as an involucre to the condensed cyme (glomerule) within. The 

 closely related C. Canadensis has a similar involucre, which, how- 

 ever, never served as scales, being enclosed with the leaves in a 

 subterranean hibernaculum, and having developed from it on the 

 approach of spring. In this it resembles many other herbaceous 

 perennials. The scales which compose the hibernaculum are al- 

 most membranaceous, and represent the entire blades, of leaves. 

 As far as can be observed from a few specimens, the growth of 

 the stem from year to year is sympodial. First a terminal bud 



