the Ophioglossaceae. 15 
ones which remained fertile. We may surmise that the next step was the 
raising of the fertile pinnae into the vertical position, so that the ancestral 
plant at this stage resembled Aneimia , as is shown in the instances of 
reversion in Botrychiam obliqunm. The next step was the fusion of the two 
fertile pinnae to form the fertile spike,and this produced the genus Botrychium, 
of which the large-leaved forms are, according to this view, the most primitive. 
Bower (6) has shown that there is a remarkable sequence of forms, which we 
may take the liberty of arranging backwards thus: Botrychium Lunaria , 
B. simplex (various forms), Ophioglossum vulgatum , O. Bergianum. Within 
the genus Ophioglossum there are signs of increasing complexity leading to 
O . palmatum ; also, as Bower suggests (p. 479), a probable line of reduction 
including O. pendulum, intermedium , simplex. On the other hand both the 
rhizome and the leaf of Helminthostachys indicate that this genus represents 
an ascending line which early branched off from the composite form which 
is believed to have resembled Botrychiitm. 
My hearty thanks are due to a number of friends who have kindly 
supplied material used in the present study: to Dr. J. C. Willis, Director of 
the Royal Botanic Garden at Peradeniya, Ceylon, for Helminthostachys ; to 
Professor E. C. Jeffrey, for a young specimen of Helminthostachys , and for 
the use of the photographic equipment of the Laboratory of Phanerogamic 
Botany at Harvard University ; to Mr. J. C. Parlin, for several species of 
Botrychium ; to Mr. F. W. Bachelder, for a generous supply of abnormal 
Botrychium obliq 7 ium ; to Professor G. L. Goodale and Mr. R. Cameron for 
many ferns from the Harvard Botanic Garden ; and to Professors A. F. 
Blakeslee and N. L. Britton, Mr. A. J. Eames, Professor J. W. Harshberger, 
Dr. O. W. Knight, Mr. C. H. Knowlton, and Mr. C. S. Ridgway for other 
material. 
Summary. 
1 . The pair of vascular bundles which supply the fertile spike in Botry¬ 
chium virginianum arise from near the two edges of a trough-shaped leaf- 
trace which is generally split into halves. Each of these bundles leaves 
a gap in the leaf-trace, which is obscured by the slight divergence of the 
bundles from the trace. Similar gaps occur in Osmundaceae, Polypodiaceae, 
and other families of ferns. 
2. The main vascular supply of the petiole runs into the sterile seg¬ 
ment, where the bundles which supply the pairs of leaflets arise in exactly 
the same manner as those of the fertile spike. It is therefore inferred that 
the fertile spike represents two fused leaflets or pinnae, namely the basal 
pair, of a fern leaf. 
3. In B. ternatum and B. obliqmim the bundles which supply the fertile 
spike arise directly from the edges of the trough-shaped leaf-trace. The 
