3 2 6 The Botanical Gazette. [December, 



the Sphagnaceae. The close similarity in the development of 

 the sporogonium in these two groups, can only be explained 

 on the assumption of a common origin, and this is strength- 

 ened by the fact that the protonema of Sphagnum is a large, 

 flat thallus instead of the filamentous form common to most 

 mosses. In short, Sphagnum forms a link between Antho- 

 ccros and the true mosses. 



This is true of the protonema, as well as the sporogonium, 

 for while the younger protonema is at first a simple flat thallus, 

 later there may grow out from its margin filaments which have 

 all the characters of the ordinary protonemal branches of the 

 higher mosses, including their peculiar oblique septa. These 

 filamentous branches, as well as the leafy stem, are second- 

 ary, and it is difficult to see how we can assume that the 

 former represents the primitive condition, as Goebel 1 and 

 Bower 2 assert. It seems much more in accordance with the 

 facts to believe that the flat thallus represents the primitive 

 form of all, and that in the mosses, as the leafy branches 

 bearing the sexual organs became more and more prominent, 

 the large prothallium-like protonema gradually became lost, 

 being replaced by the secondary filamentous form found in 

 most mosses. 



The foregoing attempt to point out the connection of the 

 different groups of the bryophytes may be illustrated by the 

 diagram on the opposite page. 



The probable connection of the Anthocerotere with the 

 Filicinere has been noted by several investigators, 3 but the 

 assumption has been usually made that the relationship must 

 be sought with the Leptosporangiatre, as these have been 

 regarded as the lowest of the forms. Of the leptosporangi- 

 ates the Hymenophyllaceae have usually been regarded as 

 the most primitive, this opinion being based mainly upon the 

 delicate character of the sporophyte. The most recent in- 

 vestigations, however, do not bear this out, and the attempt 

 of Prantl 4 to homologize the sporogonium of Anthoceros with 

 the sporophyte of Hymenophyllum involves such an amount of 

 pure speculation and so little real morphological correspond- 

 enceasto be very far from convincing 



zorg M Vo^VII^p 11 ?!" bi ° logische Studie °: Ann. du Jardin botanique de Buiten- 



* LeHgeVl B c ° ta Z V°T L "L^ "' * ^ 2 



<L c p. 62 ' P " ^^ ^ntl, Hymenophyllaceae, p. 62. 



