No. 415.] POSITION OF THE ALCA. 549 
In Cyclorrhynchus psittaculus the maxillo-palatines are nearly 
horizontal, while in Lunda they are nearly vertical, the angle for 
these bones varying between these two planes for the different 
genera, they having an obliquity of about 45° in Alca and 
Uria. The vomer also varies ; in Alca and Uria it is never 
produced as a spine in front, while this is its normal condi- 
tion in Simorhynchus and the Auklets generally. In certain 
Fraterculinze it varies between these two extremes. 
4. The mandible, though varying not a little throughout 
the group, presents in its general form the same fundamental 
characters in all. The angle is always recurved, the suran- 
gular is pierced by one large foramen, or two small ones; the 
ramal vacuity is usually closed by the splenial or dentary, and the 
sides of each ramus are more or less vertical, and the symphysis 
comparatively short. It is V-shaped when viewed from above. 
5. Asa rule, the Alce have the first pair of small free ribs 
on the fourteenth vertebra, followed by a better developed 
free pair on the fifteenth. In Brachyrhamphus and Synthilo- 
borhamphus, the first pair of free riblets may be found on the 
thirteenth vertebra. Again, the number of ribs reaching the 
sternum through costal ribs varies, the variance depending 
upon a greater number of posterior pairs in some of the 
species. As we pass from the typical auks to the puffins the 
number of pairs of ribs decrease ; for instance, Alca torda has 
nine pairs posterior to the first two free pairs, eight of which 
articulate with costal ribs, and they become long and sweeping 
behind. Lunda cirrhata, as a rule, has but seven correspond- 
ing pairs, and only six of these articulate with costal ribs; and 
they are comparatively shorter and less sweeping. 
6. The pelvis offers us no definite characters that can be 
relied upon as constant, beyond the gradual change in its form. 
It is long and narrow in Plautus, Alca, and Uria, to become shorter 
and more spreading posteriorly as we pass to other genera. 
7. The number of free caudal vertebre range from seven 
to ten, not including the pygostyle. I have found the latter 
Number in a specimen of Lunda cirrhata. 
8. When the xiphoidal extremity of the sternum is notched, 
it is 1-notched upon either side, but this is a very variable 
