Pammel — Anatomical Characters of Seeds of Leguviinosae. 105 



like the remainder of the cell-wall, brown at first, then 

 changing to a dark brown with a shade of black. The phloro- 

 glucin reaction shows that the upper light line colors rose 

 red but slightly, although the exterior walls are somewhat 

 more colored. The upper part of the Malpighian cell colors 

 pale blue, on the addition of this reagent. 



The cell-walls of the nutrient layer are lignified, not only 

 where the bundles occur but from the Malpighian cells to the 

 aleurone layer of the endosperm. With sulphuric acid and 

 iodine the cell-walls color blue. The reaction begins around 

 the cell-cavity, and then passes to the light line region, and 

 the lower cells ; the light lines appear for a considerable length 

 of time as a lighter band. Several other interesting points 

 should be mentioned in connection with these Malpighian 

 cells. The lower portion is not always blunt, but may be 

 oblique. On making a cross section and placing in chloral 

 hydrate, faint indications of cross walls in the lower part of 

 the cell can be made out. The cross walls of the cells below 

 all end at the same height. When the cells are macerated in 

 Schulze's medium the lower cells may be separated. This 

 has not been clearly indicated by some of the writers who have 

 studied the testa of this order. We may also note the occur- 

 rence of an osteosclerid layer above the Malpighian cells, 

 which supports the trichomatous epidermis. 



Geraniaceae. The following Avorks bear on this order. 

 Rober (434); Zimmermann (475); Raunkiaer (430); De 

 Toni (324); Brandza (26, 80. pi. 5.f. 5-8). 



Eober, 1877, was the first one to call attention to the light 

 line in the testa of this order. Raunkiaer, who studied the 

 development and structure of the testa in this order, speaks 

 of the occurrence of a narrow light line. The short Mal- 

 pio-hian cells are covered by small epidermal cells, and a layer 

 of parenchyma. The light line is very narrow and might 

 easily be overlooked, at least in Geranium caroUnianum. 

 Strandmark did not observe it in G. sanguineum, although he 

 shows an apparent thickening at this point. A small cell- 

 cavity occurs in the lower part of the cell. With phloroglucin 

 no reaction takes place, although the lower, light-colored, 

 elongated cells color slightly. With chlor-iodide of zinc the 



