36 Macfarlane. — Contributions to the History of 



exist also in the protoplasmic utricle, and their function is 

 probably explained by the appearance of large quantities of 

 starch during digestive secretion. 



Gardiner's "rhabdoid" has been demonstrated with great 

 clearness by directly treating surface sections with strong 

 solution of watery eosin. They should then be fixed and 

 examined in acetic solution. No evidence has been obtained 

 to support Gardiner's statement that the rhabdoid decreases 

 during secretion. A series of measurements indicate that it 

 remains the same, or slightly increases in size, but detailed 

 results will be given in a future communication. 



III. Leaf Secretion. 

 Hitherto all observers have agreed in stating that if the 

 leaf shuts through artificial mechanical stimulus, or owing to 

 the irritation of the hairs by a dry body or inorganic sub- 

 stance, what we may call "non-tetanic," closure ensues; that 

 is in such cases the marginal bristles intercross more or less, 

 but do not subsequently become everted by reflexion of the 

 leaf margin ; and that no secretion is poured out unless nitro- 

 genous matter is present. As regards the first point we have 

 already shown that it is wholly due to non-continuance of 

 stimulus, and that prolonged irritation does bring about re- 

 flexion of the leaf margins. But as it seemed possible that 

 the leaf secretion might correspond in the vegetable kingdom 

 ±o what is known as the waste metabolic material of animal 

 muscle, numerous experiments were arranged to ascertain 

 whether continued or intermittent stimuli might not cause 

 the secretion to flow. Fresh vigorous leaves on several 

 plants were carefully washed with water and left to dry for a 

 day. Small glass beads, fragments of quartz and fragments 

 of pot-crock were then laid on the leaves which were made to 

 close. A few minutes later the hairs were repeatedly irritated 

 by insertion of a blunt needle, and on examination an hour 

 after, the leaves were found to be tightly closed. They were 

 then restimulated, and the actions was repeated every two 

 hours for sixteen hours. In from eleven to fifteen hours 

 after closure the gland surfaces were moist, and after twenty 

 hours were secreting freely. 



