HEREDITY. 409 



time they are found to diverge in their evolution, start- 

 ing with the orthoceran and passing through parallel 

 lines of forms, m^ny of the genetic series having in^ 

 succession cyrtoceran, gyroceran, and even nautilian 

 forms of the fourth and fifth classes. Others are not 

 so perfectly parallel, stopping short with the cyrto- 

 ceran class of forms or the gyroceran. Many also 

 begin with cyrtoceran shells, while others diverge from 

 the gyroceran, and still other series have only nautilian 

 shells of different grades of close coiling and involu- 

 tion. 



" The application of the law of repetition in hered- 

 ity to the chambered shell-covered cephalopods, shows 

 that the straight orthoceran shells. Fig. 119, No. i, 

 were repeated in the young of the curved cyrtoceran 

 forms. Fig. iig, No. 2, and these forms in their turn 

 in the young of the gyroceran forms. Fig. iig, No. 3 ; 

 and this may be seen by comparing the young or api- 

 cal part of each shell represented in outline with the 

 full-grown shells of the preceding figures. The apex 

 of No. 2, with the whole of No. i ; the apex of No. 3, 

 with the whole of No. 2. It will be understood, of 

 course, that the figures in outline represent full-grown 

 shells, except when otherwise explained, and that they 

 were built like the shells of Nos. 1-2, by an animal 

 living in their interiors and adding band after band of 

 shelly matter to the exterior, but in these outlines the 

 shell is supposed to be perfect and the internal struc- 

 tures concealed.^ The young of Fig. 119, No. 4, 

 which represents the fourth class of forms repeats the 

 cyrtoceran form, then curves more closely, and just 

 before it comes in contact there is a short time when 



lExcept in No. g, in which a portion of the shell is broken away, showing 

 the cast of the interior and the sutures. 



