('laxf<{p''-aflon of fh<' Hrac/tlo/xxla. — Schii<-/ierf. 14;") 
In tracing the four types of pedicle openings to their origin, it 
is found that the Trhifrmidfa were the last to appear, having been 
developed in the Pfiifaiiuridd' of the Profnnuifa. The Atremata 
gave rise to the Ncntrcmota and Protrrmata. Since L'nxjuhi of 
the Atrcmafa is not the prototype for the class as it passes through 
a paterina and obolella stage, this must be looked for in a shell 
not passing through more than one stage. Paterina is this type, 
being the most primitive genus known, as well as the adult form 
I'epresenting the embryonic shell or protegulum of other brachio- 
pods. The Afrcmafa through the Trnnatid<i gave origin to the 
Neotrrmafa, while the Pnitrrmata originated in Kntoryinida., 
which is one of the flrst steps from the inarticulate towards the 
articulate forms. 
Of secondary value for classification the writer lias relied on 
the presence or absence of a straight hinge line, internal plates, 
calcareous lirachial supports, and reversional or geratologous 
development. In some families, containing chiefly rostrate 
forms, as in the Pnitanurida and N'nrlf'oyn'rida, there are genera 
with short straight hinge lines. In other families where long- 
hinges are prevalent, rostrate examples, as in the OrfJiida and 
Spirtfrrid(r^ are found. The exceptions are either specializa- 
tions or reversional tendencies, and when sufficiently pronounced 
are regarded as of subfamily importance. Examples of ger- 
atology are present in most of the four orders, but particu- 
larly in the TcrrhratclUdiv^ where the Mryatlnjrliioi and Kraiis- 
slnina: have partially or entirely lost their calcareous brachial 
appendages. 
The accompanying phvtc, (pi. v.) giving the appari'ut genesis 
of the families and their geological distribution, is added so that 
students can have before them on a single page a summary of the 
classification here proposed. It should be borne in mind, how- 
ever, that the lines are but a graphic expression of our i)resent 
information of the class, and that future study may change their 
arrangement. * 
Dair in his Index says ••from tlie [jrcecding list it appears 
that about four hundred and sixtj'-three generic and suhgeneric 
names have been rightlj' or wrongly associated with the group of 
*The names Lin(/id(ii^mi!(hr and Orthh'niidtv should be changed to 
l/iiKjidaxriKitidn jind ( 'lil(niihii/nlid;i\ 
