190 MEMOIRS NATIONAL ACADEMY OF SCIENCES, VOL. X, NO. 6. 



The water-storage-tissue reaches a high development in several of these plants. It is always 

 hypodermal and covers the entire ventral face of the leaf in Tradescantia Warszt wicziana, besides 

 that it is very amply represented on the dorsal face, but interrupted here and there by narrow 

 strata of chlorenchyma, inside the stoinatiferous strata of the epidermis. In T. rosea, Am /'/, ma, 

 and Weldenia it is distributed over the entire ventral face of the blade. It occurs as a large 

 group below the midvein in ( '. n udiflora ; above the larger veins in ( '. hirtellq ; above and below 

 the midvein in ( '. erecta and Virginica, or only above this; as in T. Virginica. 



Finally, the epidermis, witli its various kinds of projections, as warts, obtuse, pointed, or 

 curved hairs, and the stomata do also show several modifications within the family. The number 

 of subsidiary cells appears to be constant, at least in the leaves, and it deserves notice that 

 Tradescantia Floridana is the only species of the genus in which two pairs of subsidiary cells 

 were observed; all the Commetinas have two pairs, while the Tradescantim and Weldenia only one. 



We might ask now whether it is possible to classify these plants as Xerophytes, Mesophytes, 

 or Hydrophytes. Judging from the nature of the habitat where these plants occur, it appears 

 as if we might consider, for instance, T. Warszewicziana, T. sp. (from Colorado) and Weldenia 

 as xerophilous, but from higher elevations. T. rosea from low sandy woods in subtropical 

 Florida is not exactly a xerophyte in the stricter sense of the word. T. Virginica appears 

 intermediate between a mesophyte and xerophyte. Aneilema on the other hand may be well 

 classified among the hydrophytes. In regard to Commelina hirtella from sandy river-shores, it 

 seems difficult to express any opinion whether this is to be considered a hydrophyte or a 

 mesophyte. 



Corresponding difficulties are met with when we examine and compare the dominating 

 features of the leaf-structure of these same species, if these are to be brought in connection with 

 the nature of the habitat. They may he enumerated as follows: 



Aneilema: Leaf isolateral; water-storage-tissue over entire ventral face; collenchyma ami stereome on leptome-side of 



midrib. 

 ( 'ommelina hirtella: Leaf dorsiventral; water-storage-tissue above the larger veins; collenchyma on leptome-side. 

 Tradescantia Virginia:: Leaf partly dorsiventral; water-storage-tissue above midvein; collenchyma on leptome-side. 

 V. sp. (Colorado): Leaf isolateral; water-storage-tissue over entire ventral face and on leptome-side; collenchyma on 



leptome-side. 

 T. rosea: Leaf almost isolateral; water-storage-tissue over entire ventral face; collenchyma on leptome-side. 

 Weldenia: Leaf dorsiventral; water-storage-tissue over entire ventral face; collenchyma on leptome side. 

 Tradescantia Warszewicziana: Leaf dorsiventral; water-storage-tissue over both faces of the blade; stereome on leptome- 



aide and in margins. 



Of these tissues the water-storage-tissue is somewhat unequally developed so far as concerns 

 T. Virginica, since it is so very prominent in the other species. It is very scantily represented 

 in Camim Una hirtella, but, as we'remember from the foregoing, none of the other species pos- 

 sessed any large quantity of this tissue. The presence of a large water-storage-tissue in the 

 isolateral leaf of Aneilema is a frequent occurrence among hydrophilous plants. In regard to 

 the other species, the great development of this tissue may be well brought in connection with 

 the nature of the habitat. 



The collenchyma is absent from the leaf of T. Warszewicziana, but is replaced by stereome; 

 it is present in all the others, accompanied by stereome in Am il, nm. It seems strange that this 

 tissue is not developed in the leaf of T. Warszt wicziana, since it is so well represented in the 

 scape. It seems also strange that stereome occurs in Aneilt ma, but not in the species of Trades- 

 cantia from Colorado, Florida, or District of Columbia. 



The collenchymatic tissue is evidently a character of the family; the water-storage-tissue an 

 epharmonic. 



But from the present knowledge of these plants we dare not enter any further into a discus- 

 sion of their classification as members of certain associations, mesophilous, hydrophilous, or 

 xerophilous. To fully appreciate the importance of the morphological and anatomical structures 

 which we have described above it seems necessary to study many of the other representatives of 

 the family. We hope that the present investigation may prove useful to future studies as a con- 

 tribution to the knowledge of these interesting plants. 



Brookland, D. C, March, 1906. 



