INVERTEBRATA, CRYPTOGAMIA, MICROSCOPY, ETC. 485 



' Prothallium of Ferns.* — Leitgcb returns to this subject in a 

 series of investigations carried on chiefly on the prothallium of Cera- 

 topteris thalidroides ; and confirms his conclusions jireviously arrived 

 at,t viz. : — (1 j The dorsiventral structure of the prothallium of ferns 

 is an effect of light, and is no way affected by gravitation ; (2) a 

 reversal of the two sides of the prothallium may be brought about by 

 a change in the direction of the light, and the dorsiventral structure 

 is consequently not inherent in it ; and (3) the archegonia and rhi- 

 zoids are always developed on the shaded side. The jiosition of the 

 smtheridia is on the contrary very various in different species. In 

 some species they occur chiefly on the edges, in others almost exclu- 

 sively on the ventral side. 



Affinities of the Carboniferous Sigillari0e.:j:~The affinities of the 

 Sigillarice are still in dispute. English ijalfeobotanists regard them as 

 representing the highest modifications of the Lycopodiaceae. Tiie 

 French and some Americans elevate them to the Gymnosperms. The 

 only i)lants associated with the Sujillarice in the carboniferous forests, 

 that exhibit any possible affinities with them, are the Lepidodendra on 

 the one hand, and the Gymnospermous Badoxijlous and Cordaites on 

 the other. 



The old idea that Sigillarice must have consisted of lar^^e branch- 

 less stems must be abandoned. The external leaf-scars exhibit 

 nothing distinctive ; for some types have the vertical flutings of the 

 stem and the linearly disposed leaf-scars of the Syringodendrn, whilst 

 others, such as Sigillaria elegans and spinulosa, exhibit the diaf^onal 

 arrangement of the leaf-scars characteristic of the Lepidodcndra. The 

 cortical tissues in the two types of Lejndodendra and Sigillarice are 

 absolutely identical, and it is upon supposed distinctions in the vascu- 

 lar axis that French botanists rely. According to them, the Lepiclo- 

 dendra have a vascular axis on which the scalariform vessels are not 

 arranged in any radial order, nor increased in bulk by any exogenous 

 mode of growth ; in the Sigillarice, whilst the central j)art of the vas- 

 cular area is occupied by a cylinder in all respects identical witli that 

 of the Lepidodeitdra, it is surrounded by an outer zone in which the 

 vascular wedges arc radially disposed, are separated by medullary 

 rays, and have grown exogcnously through the operation of a cambium 

 layer. 



These supposed differences Professor W. C. Williamson denies 

 to be even generic, much less ordinal, since in several cases they can 

 be shown to bo duo solely to age, sj^ccimens having been found exhi- 

 biting transitional stages. When a Lcpidodendron is about to diclio- 

 tomize, the vascular cylinder splits into two horseshoe-shaped halves. 

 The same takes place in Sigillaria. The invariable dicliotomiza- 

 tion is in itself a Lycopodiaccous feature. Van Tic^'licm has 

 shown that the ultimate roots of the LycojiodiaceaB and of tlio 

 Ophioglossiaccie have a jicculiar structure of their own. In the Cycadca) 

 Conifers, and in other Vascular Cryptogams, the centre of each root 



♦ ' SB. k. Akad. Wi.ss. (Wi.ii),' Kxx. (1880), Ist Sec, p. 201. 



f Sc this Jmirnal, ii. ( I87D) |). \)\1. 



X ' IJep. 13iit. Asaoc. Ailv. S-i.,' 187!», p. .'ACt. 



