18 SCIENCE 
lariform bordered pits at the overlapping ends, 
with the exception that occasional shorter pits 
are present and in a position to form branched 
bars should the pit membranes and pit borders 
be eliminated. The lateral walls show perfect 
transition from scalariform to circular pits. 
Living types, therefore, preserve all stages in 
the transformation of scalariform tracheids 
into vessels with multiseriate circular pits 
and simple perforations; and there is every 
reason to believe that modification is still go- 
ing on. 
Many seem to be under the impression that 
scalariform pitting is of rare occurrence above 
the eyeads. A close comparison of scalariform 
tracheids, which Wieland’s material makes pos- 
sible, can leave no doubt that existing forms, 
in dicotyledons as well as in monocotyledons, 
still exhibit, in the vascular elements of their 
secondary wood, almost complete stages in the 
transformation of scalariform pitting into 
that of the circular multiseriate type, aftord- 
ing a valuable criterion by which to judge the 
relative primitiveness of angiosperm groups. 
The histologic evidence is fairly in accord 
with the floral evidence. 
The exceptional abundance of circular pits 
in such forms as Vaccinium corymbosum, 
noted by Jeffrey, is accentuated by the fact 
that, in this type, the vessels are mostly 
isolated from one another and in contact with 
wood-prosenchyma which forms circular bord- 
ered pits in common with the vessels. Sca- 
lariform pitting occurs near the pith where, 
occasionally, vessels are adjacent. Here occur 
vessels showing perfectly the transition from 
sealariform pits to scalariform perforations, 
as well as the transition stages, noted by 
Thompson,? from scalariform to simple or 
porus perforations, Forest B. H. Brown 
OsBorN BoranicaL LABORATORY, 
YALE UNIVERSITY 
THE ORIGIN OF DICOTYLS 
It is generally conceded that the origin of 
the conifers is from the evolutionary point of 
2Thompson, W. P., Jan., 1918, ‘‘Independent 
Evolution of Vessels in Gmnetales and Angio- 
sperms,’’ Bot. Gaz., LXV., pp. 89-90. 
iN. 8. Vou. XLVIII. No. 1227 
view a fairly luminous subject. Neither stem, 
leaf nor cone has wholly new or remote fea- 
tures. The group appears rather old mainly 
because so much of its history is known or 
directly inferable. Evidently, sporophyll con- 
solidation into the hard spiny unisexual cones 
mainly oceurred in the Permian, and by Juras- 
sic time was complete. There is, thus, the 
distinct connection with the dominant Cor- 
daites of the ancient world, and the cycads and 
Ginkgo amongst still existent types. More- 
over, that Ginkgo was cosmopolitan, and of 
varied form in the Rhitic further broadens the 
possibilities of conifer relationship with known 
types, while by way of seed-ferns the several 
gymnosperm lines are in common carried back 
to Pteridophytes. 
Not so the origin of dicotyls, long wholly 
without the range of scientific discussion. It 
was understood in a general way that the mono- 
cotyls did not obscure the problem. That they 
were either a contemporaneous or later devel- 
opment appeared more or less certain. Before 
the discovery of the cycadeoids, however, no 
line of attack from the viewpoint ot either 
fossil or existent evidence seemed to carry the 
dicotyls beyond the Cretaceous. Back of that 
period the decipherable record ceased entirely. 
But the discovery that the eycadeoids were a 
varied group with flowers recalling those of 
magnolias, at once suggested that the origin of 
the dicotyls could not be wholly obscure. It 
Was soon seen that the great extension of 
Magnoliaceous species in the lower Cretaceous 
must indicate a series of more primitive mag- 
nolias in the Jurassic. There might have been 
doubt as to the significance of the cyceadeoid 
flower with its pollen-bearing synangium. But 
the reduced sorus of existing cycads plainly 
stood in the complementary relationship, just 
as did the eyecadeoid microsporophyll and the 
earpellary leaf, as components of cyead fructi- 
fication. 
Nextly it was found by Nathorst and Wie- 
land that the larger proportion of Mesozoic 
cycadales were cycadeoids, in part bearing 
small and reduced types of flowers. The Wie- 
landiella of Nathorst with its slender bifur- 
