1892.] NATURAL SCIENCES OF PHILADELPHIA. 381 



Avhich is strewn with minute ovoid nests of cork cells. From the 

 well ascertained focts as to the manner of bark formation the outer 

 epiderni could not possibly have been formed in the same manner 

 as the hypoderm, namely, by the differentiation of the horizontal 

 cells, but <;ould only have originated from an independent exoge- 

 nous growth, such as a supposititious enclosure by a clasping leaf- 

 blade could produce. Aside from these considerations the contin- 

 uity of the nerves in this intra-nodal sheath Avith the nerves of the 

 leaf-blade indicated, clearly shows the identity of their origin. 



It may be said in brief that while plants, generally, in their first 

 year's growth, have only two separate systems — wood and a single 

 layer of bark — a section of Lonicera cceru/ea, and plants constructed 

 on a similar plan, have three, the outer layer of which, by the absence 

 of suber cells and other characteristics, clearly is not true bark. 

 There seems to me little doubt but square-stemmed annual growths 

 or the appearance of decurrence on growing stems, is due to the 

 fact that the leaves have really originated below the point from 

 which they seem to emerge, and that the angularity or decurrence 

 is due to the more or less imperfect meeting of the edges of these 

 leaves when clasping the stem. 



Raphanus sativus. 



The garden radish is admittedly self-fertile, but noting on the 

 14th of August an unusually large number of the cabbage butterfly 

 (Pieris rapce), as well as several other Lepidoptera, about them, I 

 was led to make an extended observation with some novel results. 

 There were a large number of plants in the row under observation, 

 and the remarkable difference in fertility in the different plants, 

 first attracted my attention. In some cases, possibly three-fourths 

 of the flowers had produced seed vessels, in others, about half, while 

 some plants had only a few scattered pods. As the plants were still 

 flowering freely, good opportunity yas afforded to see how far sexual 

 conditions misht influence these several characteristics. There were 

 found some remarkable morphological peculiarities worth noting. 



In one flower one of the basal glands had developed to a perfect 

 pistil, which was half the length of the normal one ! As there have 

 been different views of the nature of these glands, may we not regard 

 them as undeveloped axillary buds ? In this flower instead of four 

 long and two short, the whole six stamens were of equal length. 



Another flower had but three long stamens, and these a little 

 longer than the pistil at this stage of blossoming. Of course in this, 



