160 PROCEEDINGS OF THE ACADEMY OF [Feb., 



Bloomfield, Benmore woods, King' Edward Hotel woods and Cedar 

 Hill woods, many of the forms probably were driven to the hills by 

 the advancing cultivations, and the molluscan population hence 

 became more dense in these places. This would mean a mingling of 

 the lower-ground forms with the hill-top forms, yet the general hill-top 

 character is more pronounced here than in the hills at Somerset. 

 The larger forms were absorbed by the hill-top races in these cases, 

 and the clearing killed off the lower-ground forms that did not migrate, 

 so that the smaller hill-top races were no longer strengthened by the 

 constant mingling with the lower-ground forms and hence the reduced 

 size became a fixed character. These isolated hill-tops, therefore, 

 became isolated colonies, and strict equilibration of the forms within 

 contracted environment, both as to the external influences and the 

 mixture of substance due to hybridity, without mingling of the blood 

 from the larger lower-ground forms, the supply of which was now cut 

 off by their extinction due to advancing cultivation, could begin. 

 The result is the development of the small hill-top forms now char- 

 acteristic of these colonies near Mandeville, or what may be called 

 the Mandeville race of P. a. goniasmos. They differ from the lower- 

 ground forms, as has been shown, by being of smaller size and by 

 having a proportionately higher spire; but, in the colonies near 

 Mandeville, these characters are accentuated, as compared with the 

 hill-top forms found in the Somerset region, where mingling with the 

 larger, and in general flatter, forms is still possible. 



The Migrations. 



Comparing all of the forms from all of the colonies, as is done in 

 fig. 6, shows not only apparent migration lines, but indicates two waves 

 of migration into the region from the north. The first was probably 

 represented in the extinct Somerset race, and these as they moved 

 south towards Mandeville became differentiated, as I have pointed 

 out. The very old dead shells of the Mandeville region are generally 

 rather high in the spire, higher than the living forms at Somerset of 

 the same diameter. The Somerset gully forms I take to represent 

 a second migration, they are the lower-spired form. At Somerset, the 

 normal race has come from a mingling of these two stocks by long- 

 continued hybridity, and the Somerset hill-top form with its reduced 

 size is the result of this mingling. The same flat forms that resulted 

 from this hybrid stock were seen at the Kendal Road and at the 

 Somerset Road colonies; this hybrid, migrating into country occupied 

 by the older, higher-spired forms, and mingling with them, produced 



