THEORY OF RADICALS AND MORPHOLOGICAL EQUIVALENCE. 23 



clature, are epacmic, and the tertiary are what we should call acmic radicals. 

 Ccel. Pettos is an excellent example of an acmic radical in the Jura. It stands 

 morphologically and chronologically at the centre of the affinities of the group 

 of Dactyloidae and Stephanoceratidaa, that is, of the larger part of the oolitic 

 Ammonitinae. It is, in its relation to these, and to the characteristics of their 

 nealogic stages of development, an epacmic radical, but with regard to Psilo- 

 ceras, and more ancient secondary radical forms, it is a tertiary or acmic radical. 

 It has a flattened abdomen, very divergent sides, like those of Steph. coronatum, 

 and similar acmic radical forms, and a line of coarse tubercles along the sides. 

 Though altogether distinct from Psiloceras, it is also a perfectly discoidal form. 

 The direct descendants of Pdtos, which belong properly to the stephanoceran 

 and allied groups, are also discoidal forms, though the series often have inYolute 

 species, such as Meter, macrocephalum, etc. 



Tertiary radicals in what we propose to call the Pettos Stock, or Spinifera, 

 according to the evidence of the younger stages and the characteristics of adults, 

 have but one row of large spines in adults, and whorls which are very gibbous or 

 trapezoidal in section, that is, with abdomen broader than dorsum. The whorl 

 may sometimes be smooth, with only one row of lateral spines, but is usually 

 strongly pilated, the pilae being single on the sides and as a rule bifurcated only 

 on or near to the abdomen. The sutures have a more or less close resemblance 

 to those of Per. Dudressieri, or Coel. Pettos. The line of descent being broken, 

 we shall, in the imperfect list below, give some forms having two lines of tuber- 

 cles. These, however, have young which, until a late stage, show only one outer 

 line of lateral tubercles, as in the adults of the two species cited above. Steph. 

 nodosum of the Lower Oolite is the tertiary radical of that genus, and of Macro- 

 cephalites, Sphseroceras, Morphoceras, Reineckia, Cadoceras, Quenstedioceras, 

 Aspidoceras, Olcostephanus, and Pachydiscus. All of these genera have some 

 forms which are either closely similar to the radical in the adult stages, or else 

 have young with a nodosum-like stage. Peltoceras edhleta has a similar history, 

 though it is like Bactylioceras in its nealogic stages, it has two lateral rows of 

 large spines, and is similar to Asp. perarmatum in the adult. The huge coronate 

 forms of the Upper Jura, like Olcostephanus Gravesianus, etc., and the single- 

 spined forms like Aspidoceras Edwardsianus, and shells like Asp. perarmatum, 

 Rupettense, etc., with two rows of spines, are obviously in the line of stock forms. 

 In fact, one can select from the discoidal shells of these groups a more or less, 

 closely allied series of stock forms, from each of which a separate genus or series 

 of genera arose, until we find in the Cretaceous a new beginning in HopKtes 

 Royerianus and Cornuelianus for the species of that large genus, and of Acantho- 

 ceras, Pulchellia, and possibly Holcodiscus and Costidiscus. 



The cretaceous group, with nodose keels or lines of tubercles in place of a 

 keel, also belong to the Spinifera, but they form a separate phylum connected, in 

 common with such forms as Acanthoceras mammillare, with the Hoplites series, 

 and their radical is also Royerianus. The radical of Cosmoceras, Cos. Taylori of 

 the Lias, is a species with two rows of spines allied to Deroceras armatum, and 

 the adult characteristics of this species are repeated in the young stages of the 



