Mutualistie symbiosis of algw and bacteria 
with Cycas revoluta. 
ALBERT SCHNEIDER. 
WITH PLATES II AND IV. 
Recently my attention was directed to the tubercle-like 
growths on roots of Cycas revoluta. A cursory examination 
showed that they were infested by a nostoc. In my search 
for the literature on the subject I found few and incomplete 
references. Between 1870 and 1873 Reinke discovered para- 
sitic Nostocacee in species of Gunnera and Cycas. Jancz- 
ewski discovered parasitic alge in mosses, Cohn in Lemna, 
Kny in Floridee and Strasburger in Azolla. 
Reinke is to my knowledge the only person calling attention 
to an Anabaena found parasitic in a specialized parenchyma 
layer of Cycas roots. His incomplete though exact descrip- 
tion has induced me to study the subject more closely. 
Cycas root tubercles, which are simply short somewhat en- 
larged dichotomously branched rootlets, are quite common on 
most of our cultivated cycads. They occur on young as well 
as on old plants. The youngest plants at my disposal were 
about two years old. Only a few tubercles were present. A 
large, well nourished plant about twenty-four years old had many 
tubercles. They were most numerous near the surface of the 
soil; a few were wholly above and some were found a foot or 
more below the surface. Usually they are formed from the 
ends of rootlets, sometimes from the side of root branches, 
especially the single unbranched tubercles. In position they 
show evidence of negative geotropism. This is very mark 
in tubercles near the surface of the soil. Branching is always 
dichotomous (see plate II, fig. 3). Branches are short and 
somewhat spindle-shaped, the ends being bluntly rounded. 
Why they should branch dichotomously is interesting. It is 
Probably a form of atavism showing the relation of cycas to 
the vascular cryptogams. Likewise the occasional dichoto- 
mous branching of leguminous tubercles may indicate a descent 
from cryptogams.! As to color one may readily distinguish 
It may be mentioned here that the relative positions of the phloem and 
i 
xylem in the 1 ponds to that in 
vascular system of leguminous tubercles correspon Oo 
