2 



distance apart, the pistillate below and the staminate above — 

 each, except the very uppermost, subtended by a floral leaf or 

 bract As in TypJia the division is most numerous in the stami- 

 nate spike, so it is in Sparganium^ where we find only from one to 

 six pistillate heads, and from two to thirty or more staminate. 



The inflorescence of Typha is much more complicated than 

 that of any of the Aroideae, or, at least, of any of the North 

 American members of that order ; and the inflorescence of Spar- 

 ganium is simplicity itself in comparison. 



It might be suspected, perhaps, that an arrangement of this 

 type would be attended by various irregularities and monstrosi- 

 ties, and such is really the case. For one thing, the inflores- 

 cence varies indefinitely, even in the same species, in dimensions. 

 Mature spikes not over an inch or two in length are found, and 

 from this dwarfed condition they extend to a length of 12 or 14 

 inches, and the entire inflorescence varies from 5 inches to 3 feet 

 long. A similar variation occurs in the diameter of the spikes ; 

 some of them are not as thick as an ordinary goose-quill, while 

 others measure nearly an inch and a half through. 



A naked space between the pistillate and staminate spikes 

 was formerly regarded as a sufficient distinction between certain 

 species, as, for instance, between T. latifolia and T, angiistifolia, 

 but here, too, a great variation exists, so that this mark cannot 

 be relied upon at all as a ground of specific distinction. It is 

 true that Z! angtistifolia does ordinarily exhibit such an Interval, 

 and perhaps thousands of plants in the same locality will be thus 

 characterized, but in another locality the spikes will frequently be 

 continuous. The same variation occurs in T. latifolia. 



The inflorescence, also, is often interrupted in various ways. 

 Two pistillate or two staminate spikes not rarely occur on the 

 same plant, and even two of each kind on the same stem. When 

 this takes place in T. latifolia. It is generally in the staminate 

 spike, and the abnormal addition looks like a small Turk's-cap 

 set on the end. In several Instances I have seen a double 

 spiked stem in which one of the spikes occupied only one-half 

 the diameter of the rachis. Besides this irregularity, both kinds 

 of flowers will not unfrequently be mingled in patches along the 



I \ 



same spike, occurring generally in the pistillate spike. 



