425 



in the longitudinal muscles) is formed by the fusion of a 

 column of cells. This syncytium remains as the ''sarco- 

 plasm" of the future muscle. The other myoblasts then, 

 applying themselves to this column, form each a longitudinal 

 sarcostyle which, growing along the muscle, eventually 

 becomes inserted by its two extremities upon the dorsal- and 

 ventral walls of the thorax. Striations, similar to those / 

 observed in the longitudinal muscles, occur on the fourth day. 



So far as I could observe, the connection between the , 

 sarcostyles and their nuclei does not persist. The outer walls '^ 

 of myoblasts remain as the sarcolemma. 



(8) Intestinal Muscles. 



These are weakly developed ; they will be referred to more 

 conveniently in connection with the intestine. 



(9) The Muscle Insertions. 



These are entirely ectodermal cells, which during develop- 

 ment force ^side the underlying somatopleure and com- 

 municate with the developing muscles. 



Sometimes the process is quite simple. The terminal 

 myoblasts come into communication with adjacent in tegu- 

 mental cells which now support the muscles. Frequently these 

 cells chitinise entirely, or only partly, so that the muscle may 

 become inserted directlv on the hard chitinous walls of the 

 insect . 



Frequently the adjacent integumental cells elongate 

 greatiy and actively extend towards the developing muscles. 

 This may, as we have seen, give rise to the remarkable split- 

 ting up of cell columns into individual muscles, such as occurs 

 in the head muscles, leg muscles, and muscles of the ovi- 

 positor, and there can be no doubt that the integumental 

 cells are the active agents which bring this process about. 

 In the case of the dilators of the pharynx the cells of the 

 muscle insertions are formed from a much more limited area. 

 They have the same elongated appearance as have the other 

 head-muscle insertions; but there is no lateral pull as these 

 processes retract again, and the muscle constituents are not 

 pulled apart. It follows, therefore, that each pharyngeal 

 dilator muscle is the equivalent of a whole group of head or 

 leg muscles, which have all originated by the longitudinal 

 splitting of a single column. 



The insertions of the great thoracic muscles are especially 

 interesting. The myoblasts at the extremities of the muscle 

 columns approach close to the integument (fig. 135). The 

 integumental cells begin to divide, in the fresh pupa, trans- 

 versely to their length ; division is not complete ; on the other 



