64 HYATT ON THE TERTIARY SPECIES 



repeated observation, it being very important in its bearing on the law of acceler- 

 ation. Thus figs. 8, Hne Tz, 1-2, line /, pi. 4, are shells in vrhich they appear only 

 on the last part of the last whorl, fig! 3, line/, at an earlier period on this same whorl, 

 fig. 4, line/, at a much earlier stage, at least half a whorl sooner in the growth. 



In all of these the costse are closely approximate. Though there is very considerable 

 variation in this respect, the diiferences between these and var. major, as shown on 

 line / pi. 4, and on lines g, and i, where they are of about the same size, being 

 distinguishable by a practiced eye in almost all shells. 



Line g shows a variety identical with Hilgendorf's var. typica of PI. costatus, and figs. 

 1-5, line i, the equivalents of his var. plaiystomus of the same species. The costae are 

 very wide apart in the latter, which in my view are the young of the larger specimens of 

 the coarsely costate forms on line k above, while the specimens figured on line g are the 

 young of those shown on line h, in which the same peculiarities of the costa3 are observ- 

 able. The extremely uncoiled forms are in all cases, of course, regarded not as young 

 forms, but as diseased shells, which as previously observed, would of course be under- 

 sized in comparison with more healthy individuals of the same species. 



I have not succeeded in finding any hybrids or shells of an intermediate character 

 between these and the corresponding uncoiled smooth forms of PL ^SS^- The different 

 forms may be in general terms distinguished into two varieties : 1st, Shells with sharply 

 defined crowded costae. 2d, Shells with widely separated cost». These may have their 

 variations in the costae. (1), having either sharp, forward projecting, rim-like costse ; 

 (2), thick, vertical lip-like costae ; (3), thick costae overhanging, opposite to the direc- 

 tion of the growth of the shell. These modifications are due to the way in which 

 the costae are built up. In the first place, the re-building of the shell is begun on the 

 inside of the old whorl, leaving the edge of the mouth projecting forward like the 

 free edge of a frill, fig. 9, line g, pi. 4. The second is accomplished by the curious way 

 in which the new growth is begun, immediately along the flaring edge of the mouth so 

 that the resulting costae are of double thickness.^ The third is occasioned by a slight 

 overlapping of the old edge of the mouth by the new growth, so that the most abrupt 

 portions of the costae are the posterior sides, instead of the anterior, as in the first 

 instance. 



Varieties 1-2 may be subdivided in precisely the same way, but the peculiar ways in 

 which the variations are occasioned in var. 2, are more easily observed. 



Notwithstanding these facts, however, it is noticeable that the last two kinds, or abnor- 

 mal costae, are very rare in the finely costate series, and very common in the coarsely cos- 

 tate, or 2d variety. 



The distorted varieties are precisely parallel with PI. denudatus, so far as the form of 

 the whorls and the spiral is concerned, but bear the most indubitable marks of their deri- 

 vation from the various costate races above described. They are of all degrees of uncoil- 

 ing except the absolutely uncoiled, that is, one in which the extreme young is not closely 

 coiled. The whorls may not touch anywhere, after the first part of the first whorl is 

 built, but this is invariably in close contact with little bag-like ovishell, due to the prepo- 



iThis variety is figured in the coarsely costate specimen, ety, fig. 7, line i, pi. 4. 

 fig. 1, line k, pi. 4. See also for the next described vari- 



