THE SHELL. 39 



fossil Ammonites, etc., leaving only the inner or nacreous and 

 more indestructible la5'er, which thus exhibits perfectl}^ the 

 edges of the septa ; but in some cases it is only the outer la3^er 

 that has been preserved ; and frequently when the whole shell 

 has disappeared, we have perfect casts of the chambers. The 

 decomposition of the animal has contributed to form those 

 phosphates and sulphides generally present in the bod^^-chamber, 

 whilst the permeation of water deposits crystals of calcareous 

 spar on the inner walls or sometimes even fills the entire 

 chamber. Cross-sections of fossil Ammonites with the chambers 

 filled with spar, when polished, make beautiful cabinet speci- 

 mens. Sometimes, as in some of the Orthocerata, the circum- 

 jacent mud has invaded the air-chambers, but without entirely 

 filling them, because the contraction of the vascular lining has 

 left empty spaces between itself and a portion of the wall of 

 each chamber. 



Gastropoda are mostly provided with an external shell, a 

 dwelling-place and a citadel combined, the hardness and dura- 

 bilit}^ of which, as Keferstein remarks, " supplies us with the 

 best means of knowing the animal: indeed, in many cases, it is 

 the only part known, and was formall}^ the only part valued and 

 preserved in collections. Although the animal itself offers more 

 weighty and striking characters for the separation of the higher 

 groups, yet having learned the close relationship existing between 

 shell and animal, we find therein ample justification for attaching 

 especial importance to the shell in a systematic point of view." 



We have already shown how the shell is produced by the 

 mantle. 



The form of the shell is throughout regular, and is normally 

 a cone curved into a spiral, and descending in a screw-like 

 manner from the apex or initial whorl to the aperture (ii, 30). 

 Nothing can be more beautiful than the regular geometrical 

 progression of the growth of a shell or the certainty with which 

 each species and genus grows in its normal pattern, although 

 these modes vary among themselves so widely ; thus we have 

 the simple depressed cone of the Patella (Ixxxiii, 25), all aper- 

 ture and no spire, and from it every gradation, from the Haliotis 

 (Ixxxiii, 9), almost equally depressed and broad, the result, 

 however, of a very rapidly enlarging spiral, to the long, many- 

 whorled Turritella (Ixvii, 59), or theVermetus (Ixvii, 68), which 

 is a Turitella partially unrolled into a simple long tube ; — the 

 opposite of the Patella. The whorls of a spiral shell are, in most 

 cases, closely wound around its axis, and, therefore, most part 

 of their surface is in contact, each whorl partially covered and 

 concealed by its successor ; and where the axis does not lengthen 

 by the obliquity of the spiral, we find, as in the cone (Iviii, 42), 



