4i 4 Mac far lanes — Observations on Pitcher ed 
above-named form is one that has deviated least from a com- 
mon ancestral type. Thus its leaves incline to grow obliquely 
upwards, while on the one hand those of purpurea and 
S. psittacina are nearly or quite horizontal, and on the other 
those of 6 *. flava, S. Drummondii , and 6 *. rubra all grow erect, 
or nearly so. 
Next, in S. variolaris the processes from the cells of the 
conducting surface are pretty short, and both conducting and 
detentive surfaces are glandular, particularly the upper part of 
the latter. In S . psittacina, the processes are still shorter, and 
resemble closely those of S. purpurea, while a great massing of 
glands occurs at the top of the detentive surface. If we ima- 
gine the upper part of the detentive surface, in either of the two 
last, to have lost the hairs and to have retained only the closely- 
set glands, we should get the special ‘ glandular ’ surface of 
5. purpurea , and that this is a modification of the detentive 
surface appears probable when we examine hybrids of the 
species. Again, starting from variolaris , we pass to S.flava , 
with a few glands on the conducting, and still fewer on the 
detentive surface ; then to S. Drummondii , with a few on the 
upper part only of the conducting surface, and none on the de- 
tentive surface ; lastly to ,S. rubra , where both conducting and 
detentive surfaces are devoid of glands. The hairs in the three 
last likewise point to a similar deviation. That these highly- 
specialized leaves have been derived from some simpler type 
cannot be doubted, that they are all inhabitants of N. America 
points strongly to a common geographical origin, that they 
all show so many points of similarity in histology and mor- 
phology, with corresponding physiological adaptations, indi- 
cates a common line of development, and that they tend to 
vary in minor morphological and physiological details leads us 
to believe that connecting types have been lost. 
Heliamphora 1 . From the leaf-base upwards this genus 
shows externally many honey-glands, an average of 3 to the 
square mm. being encountered over the general surface, but 
1 I am indebted to Harry Veitch, Esq., for excellent fresh specimens of this 
rare plant. 
