NOTES ON THE FORMS OF CASTELA GALAPAGETA 
285 
The specimen from Chatham Island has short and robust spines, while 
the specimen from South Seymour is unarmed, as is the type specimen 
of the species collected on Chatham Island by Darwin. 
The leaf characters on this series of specimens are quite variable. 
The specimen from Abingdon Island has many spatulate leaves, and 
still others that are rather oblong elliptical, or lanceolate, figure 8. 
The most of the leaves are rather acute at both base and apex 
approaching in this respect the form duncanensis as described by 
Robinson. There is quite as much variation in the leaf characters of 
the specimens from Charles Island as there is in the manner of arming 
of the stem. The leaves are broadly spatulate for the most part on 
one of these specimens, while on the other they are much narrower, 
being lanceolate or even linear in some instances. They probably 
Fig. 8 Fig. 9 Fig. 10 
approach closely in shape the leaves on the type specimen of the 
species as described by Hook. f. 
There are many spatulate leaves on the specimen from Chatham 
Island, figure ga, but there are others which are oblong obtuse mucro- 
nate, figure gb, so that this specimen might be assigned equally well to 
either of the forms albemarlensis or carolensis on leaf characters, as it 
has the types of leaves of each. The specimen from South Seymour 
Island has spatulate leaves, figure loa, those which are lanceolate, 
figure lob, and also those which are oblong obtuse. Spatulate leaves, 
described as being typical of this form, also occur on the type specimenf 
of both the forms albemarlensis and bindloensis as has been mentioned 
above. It might be well to state in this connection that leaves of this 
shape occur to a greater or less extent on the Academy specimens from 
all of the islands where this species was collected. 
