1903] Leavitt, — Drosera intermedia 271 
involving increased complexity of certain organs, was obtained when 
the supply of nourishment to the growing points where these organs 
were in process of formation was curtailed. Reversion seemed to be 
caused by disturbance of the nutrition. 
Observations which I have made on the peculiarities of adventive 
growths of Drosera binata support the same conclusion. I have 
noted that when adventitious buds are formed on the flower scapes 
or on the roots — parts relatively large and affording abundant nour- 
ishment, especially in the case of the roots, which are stout and 
full of starch — the leaves produced are generally from the first of 
the D. binata type ; that is, not reversionary. But if the plants are 
small and appear poorly nourished they are reversionary in leaf form 
and marginal tentacles. Buds arising from the leaves, relatively 
slender parts poor in nutritives, give small plants which bring forth 
rounded leaves for a time; that is, they revert. 
In RuopoRa, ii. 149, I published notes on reversion of Berberis 
leaves. The behavior of Berberis is like that of the Drosera inter- 
media of my experiments, in that a limited food supply (in the seed- 
ling) or decreased vigor (in autumnal leaves) is associated with 
reversion to a higher structural condition. ‘The petiole is reduced to 
a mere rudiment in the full character leaf. The seedling leaf and 
oftentimes the last leaves of the season on fully matured bushes 
have not only blades entirely different from the ordinary blade, but 
in addition well developed petioles. 
When, as in the petiole of Berberis and the tentacle of Drosera 
intermedia, the structure becomes more complex, we may speak of 
the reversion as ascending. Return to a simplified state may be 
termed descending reversion. The anatomical structure of the g/and 
of the flat-headed reversionary tentacles in Drosera is simpler than 
that of the characteristic gland of the genus. In certain respects, 
therefore, the case of the Drosera intermedia upon which I have 
experimented is one of descending reversion. In other respects the 
reversion is ascending. Both sorts of reversion appear in this 
example to be occasioned by the same condition, namely limitation 
of the materials of construction. 
The occurrence of reversionary leaves of simplified type on suck- 
ers springing from the bases of tree trunks is well known. Here 
reversion would seem to be due to — it is certainly correlated with — 
an increased stock of formative materials; for such suckers com- 
