380 PROCEEDINGS OF THE ACADEMY OF [1892. 



accepted hypothesis. It seems to me that the origin of leaf-blade 

 must be at some indefiuite place below the point of departure from 

 the axis, and that "decurrence" is simply the effect of an irregular 

 meeting of the edges of the clasping leaves. In the case of plants 

 which have the young branches square or flat-stemmed, but round 

 in their after stages, it would seem that the only way of accounting 

 for it is by conceiving the union of the edges of the blades as they 

 clasp the stem. 



An examination of some specimens of Lonicera cceni/ea brought 

 me at Seal Harbor by Mr. Redlield July 22nd, clearly demon- 

 strated that this was the case. The growth of the present year is 

 square-stemmed. In cutting the stems across, mid-way between 

 the nodes, the costa of superposing leaf is clearly seen. The scales, 

 at the termination of the season's growth suggest this behavior also. 

 They are boat-shaped, evidently formed from a theoretically dilated 

 base, and meet by their edges. Later on, incipient leaf-blades may be 

 seen starting from their terminal points. That with a little modifica- 

 tion these scales could have been elongated, become united at the 

 edges, closely clasp the real bark, and then by the formation of leaf- 

 blade diverge a't the node, seems so plausible that there is little 

 doubt that this has been the process during the vigorous growing 

 season. 



I have shown elsewhere that the rifting of bark is not a mechani- 

 cal operation due to the growth expansion of the stem, but that pro- 

 vision is made in true bark for this rifting by the formation of 

 suber cells which develoj^ after a certain number of years have 

 elapsed, and which disintegrate the bark and thus permit the 

 expansion of the wood beneath. Every ligneous plant has. by a 

 specific growth of these suber cells, its own special manner of pro- 

 viding for the opening of its bark. In Lonicera coerulea and similar 

 plants with square stems, or decurrent leaves on the young growth, 

 there are none of the suber cells on the external epiderm which truly 

 formed bark should have. In the specimens brought me the "square " 

 or external layer had not been thrown ofi'', and by a little help from 

 the knife, I was able to detach the whole from a branched specimen, 

 just as if it had been a paper mask. This epiderm, having no cork 

 cells, has to burst by drying or by the expansion of the woody axis, 

 and the rupture is down the weakest line of union, namely, where 

 I have conceived the hypothetical union of the edges to have been. 

 The square stem is then changed to a round one, the epiderm of 



