When there are two or three synangia they may still be separate, but they are crowded together, and in some cases may be more or less fused. Some cases at least of irregular quinqueloctilar synangia are due to the fusion of two original synangia, owing doubtless to the close proximity of their primordia. But if originally distinct synangia may become fused in their development, we have less difficulty in understanding how the three loculi of the synangium may be due to the fusion of three primitively separate sporangia.
The most important inference from this comparison is that the repeated dichotomy of the sporophylls of the family Psilotese is an ancient feature. A real affinity with the Sphenophyllales is thereby rendered more probable.
In determining the affinities of the Phanerogams it is the custom to attach more importance to the characters of the flowers than to the vegetative characters, which are subject to many adaptive modifica- tions. Bower has recently urged* the importance of the characters of the reproductive organs, and especially of the sporangia, in determin- ing the affinities of the genera of Ferns. If we allow that the characters of the sporophylls and sporangia are entitled to more weight than vegetative characters in deciding the affinities of the Psilotese, the family must be placed in the Sphenophyllales rather than in the Lycopodiales. The whorled arrangement of the leaves of the typical family Sphenophyllese is the chief objection to this, but phyllotaxis is often a very variable character, and notably so in the Psilotese, even though it must be admitted that the arrangement in whorls appears to have been a very constant feature in the Sphenophyllea?. It would seem, therefore, that although the character of the sporophylls, and especially the sporangiophores, justifies our including the Psilotea- in the class Sphenophyllales, they yet form a family rather remote from the Sphenophyllea.
Some years ago, in 1890, Weiss and Boveri discovered excretory tubules in the pharyngeal region of Amphioxus.f Soon after Boveri
- ' Phil. Trans.,' B, vol. 192, 1900, p. 30.
t Weiss, P. E., "Excretory Tubules in Amphioxus lanceolatus," ' Quart. Jour. Micr. Sci.,' TO!. 31, 1890 ; Boveri, Th., " TJeber die Niere dcs Amphioxus," ' Sitz.-Ber. d. G-es. f. Morph. u. Phys. iu Miinchen,' Jalirg. 6, 1890.
