BUCCINUM. 261 



of small plates wliic-li are arranged in rows with the 

 plates at an angle to each other. Whilst, however, the 

 rows of the layers {h and d) are sitviated in a line at right 

 angles to the direction of the whorls, the other layer has 

 the rows running almost in the direction of the lines of 

 growth. It follows that in a longitudinal section of the 

 shell, the plates of layers (/; and d) are seen from their 

 cut faces and appear as columns, the cross striation 

 marking the cut faces of the plate. The plates of layer 

 (c) are, however, cut so that they are seen in side view, 

 and the oblique running lines mark the edges of the 

 plates. We might therefore divide the shell into two 

 layers, an outer and an inner, the latter witli three 

 subsidiary strata built up in the same way, but, as the 

 geologists would say, unconformable. We have already 

 seen that at the apex of the shell there are a number of 

 partitions cutting o^ small chambers. These are formed 

 entirely by layer [d] of the shell. 



Formation of the shell layers and 

 p e r i o s t r a c u m . — The shell is formed by the entire 

 surface of the mantle, but chiefly by the mantle edge. 

 The periostracum and the three outer shell layers are 

 formed solely by the edge, each of them farther from the 

 actual margin, whilst the innermost layer {d'j can be 

 increased in thickness throughout life by the mantle 

 immediately below it. The structure of the mantle edge, 

 with the shell secreting cells, will be given in the section 

 on the mantle. 



There is probably little doubt that the actual 

 crystallisation of the shell substance into the structure 

 seen in sections takes place outside the secreting cells, and 

 is determined to a certain extent by the constitution — a 

 mixture of conchiolin and lime — of the secretion. The 

 origin of the complex shell structures must, however, be 



