Ig10 BRIEFER ARTICLES 461 
I may perhaps be allowed, therefore, to explain in a few words how 
I tried to get at the ‘“‘Urform”’ of the liverworts. I began by saying to 
myself: liverworts must be derived from an alga-like ancestor, so let us 
see which liverwort has in its cells the greatest similarity to algae. As such 
I took Anthoceros, and started to examine it as a possible candidate for the 
position of the most primitive liverwort, but rejected it as such on account 
of its highly developed 2x generation. Having thus failed with Anthoceros, 
which I had chosen as a possible candidate on account of the cell structure 
of the x generation, I tried whether I would have better luck in examining 
Riccia as a candidate, which, on account of its having the most primitive 
2x generation, seemed to have some qualifications to fill the post of the 
most primitive liverwort. Unfortunately, I had to reject this candidate 
also, on account of the high development of its x generation. Having thus 
found that neither Anthoceros nor Riccia would fill the post, I applied to 
Sphaerocarpus as the all-round ‘“‘simplest known”’ liverwort. In this 
expression I plead guilty; it would have been better to say “the simplest 
liverwort now living.” This candidate I had to reject also, as its 2x 
generation was already too highly developed, higher in fact than that of 
Riccia. The ‘‘Urform” was therefore evidently extinct, and I concluded 
that it must have had a very simple thallus, somewhat like that of Sphaero- 
carpus, and a sporophyte somewhat like that of Riccia; for this reason, I 
designated this hypothetical form as ‘‘Sphaeroriccia.” It seems to me 
that this way of getting at the problem is not so very erratic, but fairly 
logical; but I regret if I failed to express myself with sufficient clearness; 
and I am grateful for the opportunity the review gives me to explain my 
views in this respect in a more satisfactory way.—J. P. Lotsy, Leiden, 
