7oi 
Costaria , Undaria , and Laminaria. 
are undoubtedly due to the quicker growth of the area adjacent to the ribs, 
while the latter elongate at a less speed. Yet the characteristic regularity 
of the bullating process deserves special attention. 
A similar bullation is to be seen in the young blades of Laminaria 
cichorioides and several others, and still more strikingly in the whole life of 
Kjellmaniella {Laminaria) gyrata. The bullation found in the blades of 
Laminaria bidlata and its forms comes out at a later period of life and 
is rather of a secondary nature. 
When a frond has reached a total length of 3-4 cm. another rib begins 
to appear in each marginal area. It is at first indicated by the formation 
of bullae along the external side of the lateral ribs (Fig. 11). The newly 
formed external ribs are elevated towards the ventral surface, i.e. similar to 
the midrib ; they converge downwards with the already existing ones to 
meet together at the base of the lamina. After the new ribs have been 
completed, one more row of bullations makes its appearance along the 
external side of each rib (Fig. 12). 
When the five ribs have been completed we generally find the mono- 
stromatic apex of the lamina more or less worn away. But the distromatic 
area may be still seen remaining in the upper portion of the lamina even 
after that. 
The first appearance of cryptostomata takes place in the area between 
the first three ribs, after the bullation has begun to attract our notice. The 
seats of the hairs in such an area are more or less depressed below the sur¬ 
rounding epidermis, as shown in Fig. 26, PI. LIV, but never so deeply as to 
be called pits even in an adult frond. The depression appears to be due to 
the fact that the cells in this area neither increase in diameter nor divide 
tangentially, while the surrounding epidermal cells are multiplying actively 
to extend the lamina. 
In examining the earlier stages of the cryptostomata we find a group 
of epidermal cells almost simultaneously elongated in height and each 
finally septated (Fig. 25). I could never detect a special cell which corre¬ 
sponds to the so-called initial cell commonly occurring in the cryptostomata 
of Fucaceae. 1 2 3 
The slight depression and the lack of the initial cell in the crypto¬ 
stomata of Costaria are characters parallel to those of A lariat Exact 
similarity is found in Undaria , as will be seen below (p. 705). Com¬ 
paring these members of the family with Adenocystis 3 and Saccorhiza 4 we 
might with safety come to the conclusion that the cryptostomata in 
Laminariaceous species are different from those of Fucaceous Algae in the 
mode of formation. 
1 Simons : A Morphological Study of Sargassum filipendula Bot. Gaz., vol. 49. 
2 Murray : On the cryptostomata of Adenocystis , Alaria, and Saccorhiza. Phyc, Mem., x, 1893, 
3 Ibid, 4 Ibid. 
