THE AMERICAN NATURALIST [Vol. XLIII 



frewia, however, the pith consists of only moderately 

 elongated cells which interlace somewhat as in other 

 kelps, but very much less conspicuously. The majority 

 of them are not longitudinal, but transverse in their gen- 

 eral course, so that a cross section shows more of them 

 cut lengthwise than a longitudinal (see figures, Griggs, 

 '06). Scarcely any of them are sufficiently elongated to 

 merit the name of hyphae. Very few give indications of 

 developing into trumpet hyphae. It is evident that Ren- 

 frewia presents a transition from a pithweb of simple 

 polygonal cells to the complex differentiation of the high- 

 er kelps such as Nereocystis. Such plants must of neces- 

 sity pass through the condition of Eenfrewia in order 

 to attain mature structure. We have here then a per- 

 fect illustration of the truth of His 's idea— save in one 

 respect. His contends that the developmental stages are 

 only the necessary morphological precursers of the adult. 

 But in this case they may be phylogenetic recapitulations 

 also. There is nothing in the evidence so far to prevent 

 a decision either way. 



Let us consider some other features of the develop- 

 ment. All of the young forms pass through a period 

 when the stipe is short as compared with the lamina. In 

 all which have been described above except Hedophyl- 

 lum, this condition persists until a certain very definite 

 period, after which the stipe elongates rapidly (see figures 

 of Egregia). This condition is so similar to the adult 

 stage of Renfrewia that one is tempted to consider it a 

 recapitulation of such a stage. But instead it may be 

 only a necessary physiological adaptation which the 

 young plant undergoes early in its development in 

 order to provide a large photosynthetic area to furnish 

 the food necessary for rapid growth. A priori this would 

 seem a reasonable interpretation of the facts and it may 

 be that we should consider them without other signifi- 

 S^J 1 . 5 ? ^wever, difficult to believe that the simple 

 necessary precursor of adult 

 "sia and Egregia, Eisenia and 



Renfrewioid form 

 forms so diverse as Postels 



