30 MESSRS. NEWELL ARBER AND J. PARKIN ON 
possible of their varied types of fructification, with a view to determining 
which among them exhibit features that strike us as being of а more or less 
primitive nature. Having completed this study, we have endeavoured to test 
the accuracy of our conclusions by an appeal to such evidence as is presented 
by fossil botany. The result has been that our previously ascertained notions, 
as to the constitution of the flowers of comparatively primitive members of the 
group, have been found to agree to a remarkable extent with the facts presented 
by the fructifications of some, now well-known, Mesozoie fossils. So close is 
this agreement, that the phylogeny of the Angiosperms in its broad outlines 
seems to us to be sufficiently clear to permit of the construction of a working 
hypothesis towards its solution. 
It appears to us that although the direct ancestors of the Angiosperms are as 
yet unknown in the fossil state, this line of descent ean now be traced back to 
the great group of Mesozoic Cycadophyta, and to a hypothetical race of plants 
nearly related to the Bennettitee. There would seem to be good reasons to 
connect the Cycadophyta themselves with the Fern-like Spermophytes, or 
Pteridosperms, of the Paleeozoic period, and thus the Angiosperms, on our 
hypothesis, can be derived ultimately from an as yet unknown, fern-like 
ancestor, existing at a very early geological period. 
In this connection the publication of Wieland’s full account of the 
‘American Fossil Cycads’ has furnished us with data with regard to the 
Mesozoic Cycadophyta which, until recently, we did not possess, and to the 
author of this magnificent volume we would gratefully express our indebtedness 
for the material which he has placed at our disposal. 
We have from the first recognized that what is called the problem of the 
origin of Angiosperms is in reality a plexus of problems. In addition to 
the evolution of the flower, there is the puzzling type of Angiospermous 
foliage, and many other questions to be explained in connection with this group. 
At the time when we commenced a consideration of this subject, there could 
hardly be said to be a clue to the mystery in which these matters were 
shrouded. In presenting this, the first definite hypothesis with regard to the 
phylogeny of the race, we are aware that many of the main points of our 
argument are devoid of novelty. Others, notably Hallier, have already 
brought forward arguments or facts, of which we have made free use in 
this attempt to fashion our theory. 
The subject is a large one, and the present communication is to be 
regarded as a brief résumé of a discussion which we hope to elaborate more 
completely elsewhere. For the present we have contented ourselves with a 
statement of the main features of the problem, and its solution, with brief 
evidence in support of our views. 
We would take this opportunity of expressing our sincere thanks to 
Dr. D. H. Scott, F.R.S., for many suggestions during the past three years, 
and for the interest he has taken in the progress of the work. 
