50 
grafted on the trunk ; neither has it the zone of vessels 
constant in the vegetable pith. In a transverse section 0 
a recently formed part, the pith is found to be composed 0 
irregularly sized cells quincuncially arranged ; and in a lon£ ! ' 
tudinal one, of cells very similar to those described as bein g 
found in the membrane of the crust, but smaller and whit®! - ' 
It diminishes in size with age ; in the youngest branches it 
nearly twice as large as in the tronk ; and near the root it ,s 
entirely absent; in an examination by the microscope, l* 1 ® 
newer portions are very apparently cellular, while the old® 1 ", 
parts, though of the same structure, appear as if destitute 0 
the cells, from their being so closely pressed together by lb® 
surrounding horny texture. 
The pith is the first part developed in the formation ® 
new branches. The branches are formed irregularly on 3 
parts of the axis, on the old as well the new portions, iboUr 
most abundantly on the new. The mode in which the form 3 ' 
tion goes on, is best observed in a longitudinal section. TU>* 
curious and important point I have examined in a great varies 
of specimens, old and young, in sections of ail parts and at ;il 
seasons of the year. The lirst appearance of a branch is lb® 
formation of a white speck of medullary matter, similar 1,1 
texture and appearance to the pith, and separated from lb® 
pith of the trunk by a few layers of horny fibre. At first, d‘ |S 
spot is very small, but it soou enlarges and becomes Irb* 0 ' 
gular, having its base towards the centre and its apex toward* 
the surface; it increases in size, and that portion of the a* ,s 
that lies between the apex and surface becomes less till th® 
point reaches the investing membrane; this is prolonged befo 1 '® 
it into a pointed prominence and constitutes the first outwai' 
mark of a branch, liow the first point is formed I have bee® 
unable to determine, but it is evidently the result of an acti° u 
going on in tbe axis itself; and has not the least connects 11 
with any of the accidental patches supposed by Ellis to b® 
portions of the fleshy crust. In proof of this it may b® 
observed, that in the longitudinal sections of seventeen sp®j 
cimens the bases of the pith of all the branches were situat®. 
at about the same distance from the pith of the trnnk, a» l 
the pith of all observed in the process of formation, w ‘ l 
similarly situated. Whether new branches were forming 1,1 
the old or new parts, they all began alike and passed throOS® 
the same process ; so that their formation is the result ol 8 
action going on in the axis, rather than of accident. 
In the Gorgoniadse which grow near the shores and off be 8 ^ 
lands, which beeome so much more bushy than those fr® 1 ^ 
deeper water, the branches, which are long and cross e a ® 
other, become united at their points of crossing; and in so'® 
instance* for half an inch in extent. On examination, lb 1 * 
