THE DEVELOPMENT OF THE TEETH. 163 



new cells from below. In the latest edition, however, Kolliker 

 speaks with much more hesitation on this point. 



Magitot now (1881) holds that the whole of the dentine is " a 

 product elaborated by the odontoblasts," but neither secreted by nor 

 formed by the conversion of the odontoblasts, and he denies the 

 existence of the sheaths of Neumann in toto. 



Klein believes that the odontoblast forms the matrix only, 

 whilst the dentinal fibrils are processes continued up between the 

 odontoblasts from a subjacent layer of stellate cells. 



Robin and Magitot formerly held that the dentine matrix was 

 formed by the transformation of the odontoblast cells, but that the 

 tubes were interspaces between these latter, not corresponding with 

 the axes of the cells. 



The thinnest layer of dentine, such as may be found at 

 the edges of the dentine cap, is soft and elastic, and so 

 transparent as to appear structureless. Where it has at- 

 tained a somewhat greater thickness, globules begin to 

 appear in it, which are small in the thinner, and larger in 

 the thicker portion of the dentine cap. As they are actually 

 in the substance of the cap, their growth and coalescence 

 obviously go on without any very immediate relation to 

 the cells of the pulp; in point of fact, a process strictly 

 analogous to that demonstrated by Mr. Kainie and Professor 

 Harting (see page 148), is going on. Thus, in the formation 

 of the first skin of dentine, a stage of metamorphosis pre- 

 paratory to impregnation with calcareous salts distinctly 

 precedes that full impregnation, which is marked by the 

 occurrence of globules and their subsequent coalescence. 

 The occurrence of these globular forms and consequent large 

 interglobular spaces, in the deeper parts of adult dentine, is 

 therefore an evidence of arrest of development rather than 

 of any otherwise abnormal condition. 



When the formation of the dentine and enamel has gone 

 on to the extent of the crown of the tooth having attained 

 its full length, the reproduction of new formative pulp (in 

 teeth of limited growth) takes place only over a contracted 

 area, so that a neck, and finally one or more roots are the 

 result of its conversion into tooth substance. In teeth of 

 constant growth, however, no such narrowing of the forma- 



M 2 



