Osteological Studies of the Sub-Family Ardeinec. 21 
down a pair of ribs that articulate with the sternum through the 
intervention of a pair of costal ribs. The metapophyses are short 
and stumpy, barely reaching beyond the transverse processes. 
The bone has no descending hyapophysis, though a line marks 
the longitudinal centre of the centrum below. This fades away 
gradually on the remaining vertebrae. A large pneumatic fora¬ 
men pierces the bone, on either side, behind the transverse pro¬ 
cess, and the cavities to which they lead seem to occupy all 
parts of the bone. 
In the next four vertebrae we see but little change ; they are 
all free elements ; the neural spines do not decrease any in height, 
but they become shorter from before, backwards, shortest of all 
in the twenty-third or the last free vertebra , before we reach those 
united as one bone in the pelvis. Though this “ dorsal region 
the neural canal of a Heron is strikingly small, small even in pro¬ 
portion with the size of the vertebra. The transverse processes 
are narrower antero-posteriorly as we proceed backwards, but at 
the same time reach out further from the side of the vertebra. 
As we proceed towards the pelvis we note also that the facet for 
the head of the rib gradually approaches the anterior part of tte 
centrum of each vertebra, but finally does not quite reach the an¬ 
terior margin of the side of the neural canal in the ultimate seg~ 
ment. A line joins this facet in each case, with the facet for the 
tubercle of the rib, which is at the outer posterior angle of the 
diapophysis. On either side of the beam thus formed very large 
pneumatic openings are seen in these ultimate vertebrae, and the 
trabeculae spanning the cavities within are plainly in view. 
Four pairs of haem apophyses articulate with the borders of 
the sterumu in all of the Herons that I have thus far examined ; 
the fifth pair not reaching this bone, but articulating with the 
hinder margins of the last sternal pair. The slender pair of ribs 
that claim this last pair of haemapophyses articulate with the 
twenty-fourth vertebra of which is the first one that anchyloses to 
form a part of the pelvis. 
The last two pair of vertebral ribs are without epipleural ap¬ 
pendages, and even when these processes do occur on the ribs they 
are very weak and freely articulated with the border. Herons 
have very frail ribs at the best, a fact that strikes one the moment 
we examine the thoracic skeleton of one of these birds. 
The seventeenth vertebra having a small pair of free ribs in 
