238 SHEPHERD'S P URSE. 



of two carpels laterally placed. There has been much 

 discussion concerning the cruciferous flower, chiefly as 

 to its six stamens and single whorl of four petals. The 

 most natural explanation seems to be that which 

 makes two the type number throughout, the inner 

 whorl of stamens and the single whorl of petals each 

 becoming four by chorisis. 11 The morphological sig- 

 nificance of the small glands among the stamens at the 

 base of the ovary is uncertain. 12 



The bi-carpellary ovary becomes two-celled by a 

 membranous outgrowth connecting the two opposite 

 parietal placentae. This outgrowth, not being a usual 

 part of the carpels, is considered a false or spurious 

 partition. When the fruit (a silicle) opens, the two 

 valves split away from this false partition, to which the 

 placentae and hence the seeds remain attached. 



No part of vascular plants has so constant a charac- 

 ter as the root. The root-cap and root-hairs, most 

 characteristic root structures, are much alike in all 

 cases. The primary arrangement of the tissues in 

 pteridophytes, gymnosperms, monocotyledons, and 

 dicotyledons is upon the same plan throughout. The 

 original number of xylem and phloem masses is quite 

 limited in dicotyledons, ranging from two of each (bi- 

 nary, as in Capsella) to eight, but is not constant ; while 

 in monocotyledons it is generally larger. In dicotyle- 

 dons and gymnosperms the root increases in thickness 

 by secondary growth which eventually produces great 



11 Gray, Struct. Bot., p. 206, with reference to the views of Eichler, 

 Kunth, Henslow, and others ; Strasburger, Bot. Pract , p. 587 ; Eichler, 

 Flora, 1865, p. 497, and 1869, p. 97 (both with plates) ; Bltithendia- 

 gramme, ii, p. 200, where the literature is cited. 



l * Cf. Hildebrand, Prings. Jahrb., xii, p. 10 ; Mttller, ibid, p 161. 



