Srnii, mSi?i869!] Boyal Microscopical Society, 165 
the necessary result of the perfect growth of the lobes themselves, 
or, in other words, the ends of the leaflets with which they are 
synonymous. Hence that these spaces do not denote imperfect 
nutrition and development of parts of the leaf ; neither can they be 
considered in any way as divisions taking place in a leaf otherwise 
entire, but rather as the void spaces left where the coalescence of two 
adjacent and perfect leaflets has ceased to exist. The boundaries 
being thus defined, the leaf is resolved into at least two important 
regions — the one containing an ordinary, the other an extraordinary 
venation. The structure of the entire leaf becomes thus perfectly 
intelligible. If the altered venation taking place in the commissural 
spaces is not due to anastomosis of the vessels from two adjacent 
leaflets, I would venture to ask on what other hypothesis can it be 
accounted for ? If it be allowed to be the result of such anastomosis, 
it would appear to me that a solution is afforded to the intricate 
structure of the whole of the lobed leaf. 
The leaves in the second and remaining division of lobed leaves 
are like the last in their leaflets and their mode of junction, and 
differ only in having them arranged on the stalk like those in a 
pinnate leaf. 
Among true lobed leaves the following may be cited as examples 
belonging to this second class: — Quercus rohur, Quercus ruber ^ 
Quercus lyrafa, Hydrangea quercifolia, Carduus crisjms, Cen- 
taurea coMdidissima, Cynara scolymus, &c. 
From a series of interesting metamorphoses I have selected two 
as typical of the rest, and which show the transition from the true 
pinnate to the true penni-veined lobed leaf. These are Bubus Idseus 
(Easpberry), and Juglans regia (Walnut). 
(2). The structure of the lobed leaf may be inferred from 
Synthesis, i. e. taking the leaflets of a compound leaf, and assuming 
them to have become joined at their edges, when a true lobed leaf 
would result. That this is not a mere fanciful theory in support of 
a favourite hypothesis, is shown from the fact that the leaflets do 
thus become actually joined in compound leaves. On referring to 
the leaves of the Blackberry, it will be seen that such leaves may be 
considered with equal propriety either as imperfectly formed com- 
pound leaves, or imperfectly formed lobed ones. If all their leaflets 
were separate, the leaf would be tr ue compound, — if all joined, true 
lobed. That, when the leaflets are thus united, their composite 
structure becomes identical in character with that of an analogous 
portion marked out in a lobed leaf, there can be no doubt. On this 
head the following may perhaps be considered conclusive — a kind of 
exjperimsntum crucis from which there can scarcely be any appeal. 
Take the lobed leaf of Sycamore, and the united leaflets of the 
compound leaf of Blackberry (Diagram D, Plate V.), and place 
them side by side in order to compare them. If, now, it be allowed 
