mens have very small and remotely placed spinules; the 
other extreme has stout spines up to 10 or 12 mm. in 
length. Cuatrecasas 23992, from the Rio Digua, is al- 
most devoid of spines, whereas Cuatrecasas 15031, from 
the same area, is well armed on the petioles and leaves 
with spines up to more than 1 em. in length. Hodge 
6712, from Antioquia, has remote but strong spines along 
the petioles and the midrib and, occasionally, even on the 
tertiary veins. The collection Cuwatrecasas 9604, from 
Cundinamarca, has stouter spines, even on the upper 
surface of the leaves. If we can judge from the admit- 
tedly limited material at hand, we may suggest that there 
is evident a perceptible increase in density and size of 
spines as one proceeds northwards. Some of the material 
from Cauca is only weakly armed; the stoutest armature 
is found on specimens from the central and northern 
Andes of Colombia and Venezuela. Collections from in- 
termediate regions, such as Cuatrecasas 22694, from the 
Departamento del Valle, would seem to be links between 
Solanum quitoense and its var. septentrionale. 
The recognized fact that the density and size of the 
spines vary so much does not, we feel, argue against the 
separation as a distinct variety of the spiny from the 
unarmed variants. The genus Solanum tends to be ex- 
tremely variable in respect to spines where they occur. 
There is a very distinct possibility that, when ample 
material is available for study and when adequate field 
studies have been carried out, the variation in color and 
density of the soft indumentum of the leaves may also 
be found to be sufficiently important to use in the recog- 
nition of additional varieties. However, we cannot, at 
the present state of our knowledge, evaluate the charac- 
ters which may reside in the differences of leaf pilosity 
in the Solanum quitoense complex. 
We have chosen the varietal epithet septentrionale to 
[ 102 ] 
