160 MINNESOTA BOTANICAL STUDIES. 



fifty per cent, alcohol, and after staining, mounted in glycerine 

 jelly. Free-hand sections were found best for the cross-section 

 of the stipe. Bismarck brown, iron-alum-hsematoxylin, and 

 Delafield's hasmatoxylin, were on the whole the most satisfactory 

 stains, although aniline blue and safranin were found useful for 

 sections of the stipe. 



The Holdfast. The holdfast area consists originally of a 

 disc-like structure, as in Nereocystis and other kelps. In the 

 smallest specimen at the writer's disposal the first few hapteric 

 branches had already appeared but the primitive disc could still 

 be made out without much difficulty. The hapteres originate 

 in a special growth region at the base of the stipe (jig- /<?, fig. 

 77, , fig. 26) just above the primitive disc. They branch 

 dichotomously. In cross-section (fig. 77, b) it is seen that no 

 pith area is present in the hapteres, but that their central tissue 

 is parenchymatous, consisting of rather large cells of irregular 

 shape, which gradually become smaller as the hypodermis and 

 epidermis are reached. Chloroplasts are very common in the 

 hypodermal tissue of hapteres, especially the upper portion. 

 By comparing the longitudinal and cross-sections of the haptere, 

 very little difference between them is found (figs. 18 and 7p). 

 Growth rings, such as are reported for Pterygophora^ were not 

 found in this species of Alaria. 



The Stipe. To study the stipe the smallest specimens col- 

 lected were first used, then some of the larger plants on which 

 gonidiophylls were just beginning to appear, and finally, 

 mature plants. 



Mucilage ducts are not present in Alaria. In general, three 

 areas of tissue may be made out, vtz. 9 epidermis, cortex, and 

 pith, which latter is of the same elliptical shape as in the other 

 kelps (fig. 77, <z, b). It is chiefly composed of an inter- 

 woven mass of anastomosing filaments, embedded in a gelatin- 

 ous matrix (fig. 75). The cortical area may be divided into an 

 outer and inner layer. The cells of the inner cortex tend rather 

 towards a round shape than a hexagonal (fig. 14), but as the 

 outer cortex is reached the cells become polyhedral in shape 

 and gradually smaller in size (Jig. 13). The epidermal cells 

 appear rectangular in shape. Chloroplasts were found only in 

 the outermost layers of cortex. In a cross-section of a mature 

 stipe, the cells of the outer cortex give the appearance of being 

 radially arranged. This is no doubt due to the radial division 



