a 
ioe 
THE STRUCTURES OF PITCHER PLANTS. 297 
from S. purpurea have failed. Fluid taken from pitchers containing 
insects generally gives a distinctly alkaline reaction, and in the few 
“instances where I have obtained a faintly acid reaction, it has seemed to 
me to be due to the presence of ants. In virgin pitchers, fed with 
albumen, no acid reaction has ever been obtained, and the albumen 
rapidly decomposes. I may here state that the ordinary method of 
testing these fluids by litmus paper is crude enough not to be always 
trustworthy. A more delicate way is to make a concentrated solution of 
litmus in distilled water, and add to it about ten per cent. of absolute 
alcohol. A drop of this should be placed on a white porcelain plate, 
side by side with a drop of the fiuid to be tested, and the drops then 
made to touch. An amount of acid may thus be detected which will 
escape the eye of the observer if litmus paper be used, and there can be 
no possibility of error. I draw the conclusion, therefore, that the 
glandular structure of S. purpurea is purely absorbent, that its pitchers 
are merely passive insect traps, and that the advantage gained for the 
plant by the destruction of flies is to be attributed entirely to their 
maceration. Another argument in favour of this, the importance of 
which will be seen by and by, is that flies continue to live an indefinite 
time after having been introduced into a Sarracenia pitcher containing 
fluid. The very interesting observation of Prof. Riley concerning the 
habits of the Sarcophaga sarracenie would show that considerable 
advantageis gained for the plant by the direct application of the insect 
debris to the roots of the plant. I am quite certain from my experi- 
ments with the nutrition of Nepenthes that if S. purpurea had a secre- 
tion at all like it, no insect could visit the latter with impunity. A series 
of experiments made during the summer with test tubes of various 
sizes and diameters, and containing fluids of various kinds, have con- 
vinced me that as far as the common house fly is concerned, no specially 
disguised or attractive form of trap is required. But there is no doubt 
that the addition of the coloured venation on the lip of the S. purpurea 
must make it more attractive to certain kinds of insects, as Sir John 
. Lubbock has shown that bees are greatly influenced by colour. The 
armatures of the upper zones must also be advantageous by imprisoning 
the insects. It will be seen, then, that I differ from Dr. Hooker in that 
I regard the first and second zones of the S. purpurea as the truly deten- 
tive surfaces, and the third and fourth as absorbent. 
S. flava.—l examined the pitchers of a young plant in which the lip 
covered the mouth of the pitcher. Ii I may argue from the facts 
observed in Nepenthes this ought to be the most active condition of the 
pitcher. I found spiral vessels in the lid, and numerous stomata and 
ostioles. The latter in this case suggested that they may be the local 
centres for the growth of epithelium, for the cells in their immediate 
neighbourhood were all small and seemed to radiate from the ostiole as 
from a centre, and the intercellular canals seemed to grow with them 
(see Fig. 7, also Fig. 4 from S. rubra.) There were also a few multifid 
buds. The first and second zones of this pitcher resemble those of 
S. purpurea. The third has short tubular trichomes and no glands, and 
the fourth has long tubular trichomes. This variation in the third zone 
is noteworthy. I found no evidence of secretion here, and when the 
pitcher was over-fed, whether by albumen or naturally by a too large 
fly, the decomposition spread to the parenchyma of the leaf and killed it. 
[TO BE CONTINUED. | 
SS 
