THE HUMAN STERNUM 27 



regularity In the mode of fusion of the parts of the mesosternum (see 

 Table VIII). There is a tendency for the parts to unite together from 

 below upwards, and only exceptionally in the opposite direction. Precise 

 data are not possible about the fourth element, as it is so often absent ; 

 when present, it is usually the first to unite with the preceding portion. 



Between the time of birth and the end of the tenth year (Table VIII) 

 three or four pieces of the mesosternum remain separate. Between the 

 eleventh and fifteenth years inclusive the third (or the fused third and fourth) 

 segment begins to unite with the second. Between sixteen and twenty 

 years inclusive the second, third, and fourth pieces have usually united, and 

 the second is beginning to join the first. Between twenty-one and twenty- 

 five years all the elements become fused, so as to form a single bone. 



Eight anomalous cases have been noted, in which union occurred in 

 different ways (Table VIII). These anomalies are all traceable to the 

 behaviour of the second piece of the mesosternum, which fuses with the 

 first or third piece or with both, leaving the other element or elements 

 separate. 



These observations confirm and amplify the statement of Dwight"'"' 

 and others regarding the date of completion of ossification of the meso- 

 sternum. They Indicate, In my opinion, a greater degree of precision than 

 Dwight's cases admitted. In the process of union of the several elements 

 (PL V, Fig. 37). 



In examining the ossification of the mesosternum it is difl'icult in some 

 cases to decide on the relation to particular ribs of a given centre of ossifica- 

 tion, and one has to admit the general conclusion that the osseous deposit 

 occurs with characteristic irregularity. It appears to be stretching the facts 

 observed too far to assume that a given centre is actually associated with 

 any given pair of ribs. At the same time, an Important point Is demon- 

 strable, that even In the most regular and symmetrical cases there are only 

 three, or at the most four, sets of centres for the mesosternum. No doubt 

 five or more centres may occur, but I have searched in vain among the examples 

 examined for the presence of a distinctly segmental fifth mesosternal element, 

 such as has been figured and described by Otto.'° One does not doubt its 

 exisjtence, but one fails to give it the morphological Importance ascribed to 



