200 THE JOURNAL OF BOTANY 



in fact, the prothallus of a common Fern, and .still more that of Khizo- 

 carps, may show how precociously and efficiently sexual mechanism 

 may he associated with limiting terms of early embryology in a soma, 

 itself the limiting decadent expression of the morphological deteriora- 

 tion of a high-grade plant-form with the stem, leaf, and root of the 

 sporophyte-phase l . 



Having accepted the connection of the ascus of the Laboul- 

 beniacese with the asexual phase of the general Ascomycete series, 

 and demonstrated the identity of types of Laboulhenian somata, and 

 their relation to early stages of Pre-Floridean or Para-Phaeophycean 

 marine origin, it is now interesting to compare the details of the 

 sexual processes with those of the Florideue on the one hand, and 

 with those of Lichen Fungi on the other ; since spermatogamy is not 

 only apparently still dominant in the group, but it is specialized to 

 an extent so much in advance of the Floridese that it may cover the 

 still somewhat vague details of the Lichen-mechanism. As already 

 noted, the asexual generation is reduced to the limit of a few asci 

 contained within a ' perithecial ' investment of gametophyte tissue, 

 constructed usually of a minimum number of cell-units, functional 

 before fertilization, and controlling the protection, nutrition, and 

 spore-output of the ascosporophyte by a narrow ostiole, exactly in the 

 manner of the cellular procarpial branch of Polysiphonia, rather than 

 that of a Pyrenomycete hyphal perithecium. It thus presents the 

 morphological expression of a distinct ' procarpial ' branch of the 

 parental soma, often markedly differentiated from the somatic raniuli 

 when the thallus retains any definite mass or cellular structure 

 (Zodiomi/ces). While the asci are normal for a Pre-Floridean type of 

 unilocular sporangium, special interest centres in the sexual organs 

 and mechanism, so far unique in the plant-kingdom, though com- 

 parable and clearly analogous with conditions of spermatogamy in 

 both Lichen Fungi and Florideae, and obviously the outcome of similar 

 environment, even if always divergent in details of morphological 

 elaboration. Spermatogamy attains its limiting character ; fertili- 

 zation is restricted to an aqueous medium ; but a drop may suffice, 

 and the minimum plankton-rate of a drop, as opposed to the reef-pool 

 habit of the Floridean, bears comparison with the antherozoid out- 

 put of a Fern-prothallus, giving large motile flagellated zo'ids at an 

 effective plankton-rate in the plankton-phase of a dew-drop. Spermatia 

 are budded off in profusion from antheridial ramuli, as tufted sub- 

 dichotomous growths 2 , in the form of conidio-spermatia 3 , or again 

 ejected as minute apparently naked protoplasts within a protective 

 tube-membrane. In extreme cases special protective 'antheridial re- 

 ceptacles ' 4 express a xerophytic adaptation, fundamentally analogous 

 with the antheridial receptacle of the transmigrant Chara 5 . In the 

 first case the Laboulbeniacea? resemble the Lichen-type; in the last 

 they are more advanced than the most specialized Floridean 6 . Since 



1 ' Thalassiophyta,' Bot. Mem. 3, p. 81. 



2 Thaxter (1896), t. 2, fig. 7. 

 A Loc. cit. t. 23, Zodiomyces. 



4 Lor. cit. t. 4, fig. 16 ; t. 7, figs. 9, 22. 



5 'Thalassiophyta,' Bot. Mem. 3, p. 15. 



* Cf. Yamanouehi lor Polysiphonia, Bot. Gazette (190C><. 12. p. H" ; also 

 Graciliaria, 



