38 Rhodora [Marcu 
to be found in the rosette of leaves of the first year. As I was col- 
lecting Arabis laevigata for comparison I found growing with it young 
plants of such a different aspect that I could not at first believe that 
they were the seedlings of this species, but further observation con- 
vinced me that they could be nothing else. These young plants in 
May and June of the first year produce spatulate to broadly obovate 
leaves, sometimes reaching 8 cm. long by 4 cm. broad, which are 
sparsely to somewhat thickly pilose, especially above, with very short 
stiff hairs. This character seems to have been overlooked for the 
most part as all the later descriptions which I have consulted as well 
as most of the earlier ones speak of Arabis laevigata as entirely glabrous. 
However, DeCandolle in his Systema Naturae says: — “Folia radi- 
calia .... superne pilis rigidulis sparsis scabra," and, as will be seen 
later, the type-specimen shows just this character. Later in the 
season the leaves are smaller and more spatulate with fewer hairs, 
but I have yet to see a specimen of Arabis laevigata with the leaves 
of the first year preserved which does not show upon them some of 
these characteristic hairs. 
In the new plant the leaves of the first year are all spatulate to 
narrowly obovate and perfectly smooth and glabrous. They are 
much more persistent than those of A. laevigata, being usually well 
preserved at flowering time and consequently are usually found on 
herbarium specimens. The flowering season of this plant is two to 
three weeks later than that of A. laevigata. The latter comes into 
flower in Connecticut about May Ist, or, in exceptional seasons, as 
early as April 15th and ceases flowering about June 1st. The other 
I have not been able to find in flower before May 15th and it is in 
full flower nearly or quite throughout the month of June. 
After the observations noted above a question arose as to the 
identity of the original Turritis laevigatus Muhl. This seemed fairly 
plain from the description of Willdenow but to make the matter 
more certain Prof. M. L. Fernald kindly obtained for me a tracing 
of the type-specimen from Berlin. This shows a rather poor specimen, 
but one, which, with its sagittate-clasping stem-leaves, all undivided, 
and its obovate pilose radical leaves of the first year, is manifestly 
the same as the commoner plant which has been familiar as Arabis 
laevigata. I propose the other plant discussed above as 
ARABIS viridis n. sp. Planta 2-6 dm. alta, simplex vel paulo ramosa, 
foliosa, viridis, glabra; foliis radicalibus lanceolato-spatulatis, dentatis 
