October 13, 1911] 



SCIENCE 



All 



close that he regards them as Protocy- 

 cadeffi, the fern-like habit being restricted 

 to the position of the sporangia on the 

 vegetative fronds. Medullesa, therefore, 

 would be only one link in the chain con- 

 necting the Cycads with the Filicales, and 

 a link very near the Cycadian end of that 

 chain. Other forms more closely connected 

 with the Pilieinean phylum are still to be 

 sought. 



In bringing Professor Chodat's views to 

 your notice, I do not wish to urge their 

 acceptance, but his criticism seems to me 

 sufficiently weighty to demand a careful 

 reconsideration of the structure and affini- 

 ties of the Lyginodendrese, which, whatever 

 may be their ultimate position in our 

 scheme of classification, will continue in 

 the future, as they have done in the past, 

 to command the attention of all botanists 

 interested in the evolution of plant life. 



If the whole-heartedness with which we 

 in England received the theory of the 

 Cycadian affinity of Lyginodendron has 

 laid us open to friendly criticism, I am 

 afraid some of us may be accused of ex- 

 ceeding the speed-limit in our rapid accept- 

 ance of the Cycadoidean ancestry of the 

 Angiosperms. Ever since Wieland put 

 forth the suggestion in his elaborate mono- 

 graph of the "American Fossil Cycads" 

 that "further reduction and specialization 

 of parts in some such generalized type, like 

 the bisporangiate strobilus of Cycadoidea, 

 could result in a bisexual angiospermous 

 flower," speculation as to the steps by 

 which the evolution might have been 

 brought about has been rife, and Hallier 

 in Germany and Arber and Parkin in 

 England have put forward definite schemes 

 giving probable lines of descent. Arber 

 and Parkin in their criticism and detailed 

 suggestions connect phylogenetically with 

 the Bennettitales, the Ranales, as primitive 

 Angiosperms, and displace from this posi- 



tion the Amentales and Piperales, which 

 were regarded by Engler as probably more 

 closely related to the Proangiosperms. Of 

 course, the resemblance between the am- 

 phisporangiate, or, as I should prefer to 

 call it, the heterosporangiate "strobilus" 

 of Cycadoidea, and the flower, say, of Mag- 

 nolia is very striking, and the knowledge 

 we have gained of the structure and organ- 

 ization of the Bennettitales certainly in- 

 vites the belief in a possible descent of the 

 Angiosperms from this branch of the great 

 Cycadian plexus; but the ease with which 

 the flower of the Ranales can in some re- 

 spects be fitted on to the "flower" of Cy^ 

 cadoidea raises suspicion. Critics of the 

 Arber-Parkin hypothesis may possibly in- 

 cline to the view that "truth is often 

 stranger than fiction," and that the real 

 descent of the Angiosperms may have been 

 much less direct than that put forward in 

 these recent hypotheses. The particular 

 view of the morphology of the intraseminal 

 scales and seed pedicles adopted by Arber 

 and Parkin is, as they admit, not the only 

 interpretation that can be put upon these 

 structures, and the views on this point will 

 probably remain as various as are those of 

 the female cone of Pinus. Even if we re- 

 gard the ovulate portion of the Cycad- 

 oidean "flower" as a gynecium, and not 

 as an inflorescence, we are bound to admit, 

 as do Arber and Parkin, that it is highly 

 modified from the pro-anthostrobilus type 

 with a series of carpels bearing marginal 

 ovules. Cycadoidea was evidently a highly 

 specialized form, and may well have been 

 the last stage in a series of extinct plants. 

 Arber 's very sharp separation of mono- 

 and amphisporangiate Pteridosperms does 

 not seem to me quite justified. Amphi- 

 sporangiate forms may have been pre- 

 served, or may have arisen anew in various 

 groups of Pteridosperms or in their de- 

 scendants. Heterospory, we know, orig- 



