RAYMOND: NEW FOSSIL STARFISH. 167 
however, and Schuchert has already pointed out a number of 
ways in which the type differs from Palaeaster incomptus. 
In regard to the ambulacrals of Hudsonaster, Schuchert 
(loc. cit., p. 55) says in the generic description: "The ambula- 
craha of each column may be so closely adjoining as to leave 
almost no space for the podial openings, or there may be later- 
ally between adjoining ossicles very large openings." In any 
investigation of the ancestry of the Asterozoa, the develop- 
ment of pores between the ambulacral plates is one of the 
dominant questions, and although it is possible to group in one 
genus species showing the two conditions mentioned above, it. 
seems somewhat forced to associate so closely such extremes in 
this respect as Protopalaeaster narrawayi and Palaeaster matu- 
tinus. I would, therefore, prefer to see Protopalaeaster returned 
to its generic rank, although I do not, of course, concur in Hud- 
son's original interpretation of the specimen which served him 
as the type. 
As a matter of fact, the so-called podial openings of neither 
Palaeaster matutinus nor any other pre-Devonian starfish now 
known entirely perforate the ambulacral plates, but are merely 
excavations in their lower surfaces. Ampullae may have been 
located in these cavities, and eventually, by a continued thin- 
ning of the plates above them, have worked their way within 
the cavity of the arm. 
The process of elimination leaves three other species referred 
to Hudsonaster by Schuchert. These are Hudsonaster milleri 
Schuchert, Palaeaster incomptus Meek, and P. matutinus Hall. 
The first of these is known only from an imperfect specimen 
but the two others are represented by fairly complete material. 
The most striking feature of P. matutinus is the series of ambulac- 
rals which because of the large size of the podial excavations 
are reduced on the ventral side to almost linear dimensions. 
For Hudsonasteridac of this type, with excavated ambulacral 
ossicles, and with approximately the same number of ambu- 
lacral as adambulacral plates, I pro]iosc the generic name 
Macroporaster with Palaeaster matutinus liall as the type. 
Since Hall's original specimens did not show the actinal surface 
I designate as the genocotypes the specimens figured by Schu- 
chert in his figure 2, Plate 2 (No. 2G in Museum of Comparative 
