BOT.— Vol. I.l CAMPBELL— SPARC ANIUM. 32 1 



VIII. The Affinities of Sparganium. 



The genus Sparganmm has until recently been associated 

 with Typha in the family Typhaceae, but Engler (1889) has 

 pointed out that the differences between Tyfha and Spar- 

 ganium are so great as to make such a union unwarranted, 

 and he proposes placing S^arganmm in a separate family, 

 following the Pandanacese. Until more is known about the 

 development of the Pandanacese, however, it will be impos- 

 sible to decide how closely the two families are related. A 

 thorough study of some species of Pandaniis would be of 

 great value in solving this question. 



Hegelmaier (1874) has referred to certain resemblances 

 in the embryo of Sparganmm to that of the Gramineas, and 

 my own study of S. simplex confirms this. The extraor- 

 dinary character of the antipodal cells also is strongly sug- 

 gestive of the grasses. The probably terminal origin of the 

 single ovule is also like that of the grasses, and this it also 

 shares with most other low Monocotyledons. Indeed the 

 solitary uni-ovulate carpel is probably the most primitive 

 type among the Monocotyledons, and is not the result of a 

 reduction. In this particular S. simplex is probably more 

 primitive than the forms like S. Greenii and S. eurycarpum, 

 in which the ovary is compound. 



The type of fruit in Sparganium is not unlike that of the 

 Gramineae. While the majority of the latter have a cary- 

 opsis, there are others in which the fruit is nut-like, corre- 

 sponding to that of Sparganium (see Hackel, 1890). A 

 comparison of some of these forms with Sparganium would 

 be interesting. 



The monoecious flowers of Sparganium also are shared by 

 some grasses, e. g. Zea, so that the possibility of an actual 

 relation between these two puzzling families may very well 

 be considered. Of course the resemblances may be purely 

 fortuitous, but it is more likely that they indicate a real, 

 even if remote, affinity, and any further study of the flower 

 and fruit of the Gramineae should consider this. In short, 

 so far as a comparative study has been made, Sparganium 

 shows a closer affinity with the Gramineae than with any 

 other family, unless possibly the Pandanaceee. 



