620 



THE AMERICAN NATURALIST [Vol. XLT 



the summit of the ambital angulation. Tlie second row of spines 

 generally forms on the spiral which separates the body of the whorl 

 from the spindle. The larger spines are formed on the primary 

 spirals, the smaller intermediate ones on the intercalated secondary 

 and tertiary s[)irals. In a number of types these spines are inde- 

 pendent {Murex temiispina) while in others they become com- 

 pounded. Thus in the majority of Murices, the spines of the later 

 cycles do not become independent of those of the earlier cycles, 

 since these earlier ones grow to such an extent that the independence 

 of the smaller spines is impossible. Hence they become modifi- 

 cations upon the sides of the larger s])ines, and thus is produced 

 the wonderful complexity of spine in such forms as Murex 

 palviarosae and others. Here, too, as has already been shown 

 (1 , p. 934) the increase in complexity is progressive from period to 

 period, the first formed spine on the ambitus (or shoulder angle) 

 always leading the others. The degree of advance of the first 

 over the second spine varies in amount in differently accelerated 

 types; this variation is, however, phyletic and as such also progres- 

 sive. In some types the second spine always has the complexity of 

 the first spine in the preceding period, in others its complexity is 

 comj)aiahle only with that of the primary .spine in a still earlier 

 period. 



( 'nlumr/lar plivatlous and lirac—Vhe developuKMit of these feat- 

 nrv. alx. fc.llou. the lau <,f pio-rc.Mxc a;)p(>arancc and progres- 

 sive intensification. Those tvp.N hkc I-;H<ioliina, Khopalithes, 

 Latirus, etc., which in the ;i(hilt have M'vcral plications, arc as a 

 rule, free from the.se in the carli(\si stages. 'V\\v onlv exception is 

 in the case of accelerated types, which start conch-life already 

 equip[)ed with plications, as many of tliem start equipped with 

 costae and spirals. In the more primitive types, however, the 

 plications appear in succession, the first while the shell is still 

 young, the others succes^iv <1\ . In >oinc ca.ses the second ])lica- 

 tion has only reached halt the m/.c oI the first in the adult shell. 

 In types w^th niiineron> cohnneliar i)licati()ns, intercalated ones 

 appear bet we<'t I tlu^ ol<ler one-. Tlie-e Keuin chiefl> in later stages 



