22 2 ^ The Ohio Naturalist. [Vol. XI. No. 2, 



The origin of the primary shallow furrows is different from any- 

 thing; seen in either of the preceding genera. In Lessoniopsis the 

 relative increase in anticlinal activity in the epidermis seems to be 

 the factor operative in reducing the thickness of the cortical layers. 

 The undue stress brought to bear on the cortex by the rapidly 

 expanding superficial layer results in the premature transition 

 of the inner and middle cortex to the pith-wT'b condition. Ordi- 

 narily expansion at the surface in inanimate objects results in 

 buckling. In this case the transmission of the stress to the inner 

 cortical layers stretches their elements into the thinner or highly 

 elongated ]Dith-wcb condition, thus markedly lowering the upper 

 layers so as to produce concavity instead of convexity at the sur- 

 face. No figure is given to illustrate this condition as the area 

 concerned was far too extensive to be drawn on a scale sufihciently 

 large to show the histological changes. 



The second stage or splitting proper is by a process as different 

 from that observed in Postelsia as that is different from Nereocys- 

 tis. The central part of the mid-rib, after its reduction to the 

 thickness of ordinar>' lamina, is locally still further reduced by 

 the action of an internal cortical meristem, associated with a 

 quiescent epidermis. On both sides of the medulla in the region 

 concerned the cortical cells are seen dividing with anticlinal walls 

 (Fig. 12). This has resulted not only in severely attenuating the 

 pith-web but on one side the epidennis itself has parted and the 

 critical point has been reached, for the thin-walled cells of this 

 newly foniied tissue cannot withstand the ripping tendencies in 

 the wave swayed and twisted lamina. At approximately the same 

 stage or on the same slide from which the drawing was made, the 

 lamina portions were already separated. 



In healing, the wounds first are covered for a time with a callus 

 formation but later the epidermis and cortex heal them o\'er 

 exactly as in Nereocystis. 



MACROCYSTIS. 



The splitting process in this genus was studied by the Gcmian 

 writers already quoted, but they evidently confined their attention 

 to the development of the original perforation and did not study 

 the elongation of the cleft, which is carried out by a dififerent pro- 

 cess than that forming the ]jerf oration. This fact at once sets 

 Macrocystis apart from the preceding genera in which the processes 

 originating the perforation are also operative in elongating the 

 cleft. Sufficient material was available showing the incipient and 

 older splits, to make four or five series of sections illustrating each 

 of these stages. Part of the material studied came from the 

 • ruvian coast, part from Vancouver's Island. Sinall and large 

 venilc laminae, having splits in about the same stage of develop- 



