On the structure of cross-striated muscle. 201 



lemmal protoplasm, but on this point I have no evidence to öfter, as 

 I was not successful in getting any such fibres to stain in this way. 

 It is impossible to believe, with the evidence of such preparations as 

 these before us. that the fibres of these transverse networks are of exactly 

 the same nature as the septal portions of sarcoplasm which give rise 

 to the appearance of the longitudinal lines. It is true that in ordinary 

 gold-acid preparations both become stained. But when, as in the case 

 here described, the access of the gold is restricted, the transverse 

 fibres evidently exhibit a much greater affinity for the metal. 



A further proof, if one were needed, of the difference between 

 these filaments of the transverse networks and the rest of the sarco- 

 plasm may be obtained from their appearance in the living fibre. 

 Here they appear in section as rows of dark, round dots (Figs. 1 

 and 5), and exhibit a much higher refrangibility than the rest of the 

 fibre. Indeed, their effect here upon the light transmitted through the 

 fibre is such that each dot appears surrounded by a bright halo, 

 which is unquestionably due to the presence of the dot 1 ) (or rather 

 to the filament of which the dot is the optical section) for the haloes 



*) This fact, which was demonstrated in my paper on the leg-muscles of Dytiscus 

 in the Philosophical Transactions for 1873 (p. 433) but the significance of which 

 has escaped most subsequent writers, has been re-affirmed by van Gebuchten (La 

 Cellule, T. II, p. 392, 1886). Were the filaments merely extra-accumulations of the 

 same interstitial substance as occurs elsewhere between the sarcostyles, and nothing 

 more, it is difficult to understand why they should appear so perfectly round: one 

 would expect them to appear angular. At the same time they have evidently much 

 in common with the rest of the sarcoplasm, for the latter also has a considerable 

 affinity for the metal and they are both stained, in acid preparations, by safranin. 

 Moreover the transverse networks become in contraction increased in bulk, ap- 

 parently at the expense of the remaining sarcoplasm. It is therefore still possible 

 that Rollett's idea that the whole of the sarcoplasm, including the transverse net- 

 works, represents the remains of the original protoplasm of the muscle-cell, may be 

 the true one; but it must I think, be modified by supposing that the granular part 

 of this protoplasm has remained for the most part around the nuclei, and in the 

 position of the transverse networks, whereas the clear or homogeneous part occupies 

 the remaining interstices between the sarcostyles. This is the case in most muscles, 

 but in some the intercolumnar sarcoplasm is throughout granular. That cell-proto- 

 plasm is thus capable of becoming differentiated into these two parts, is shown by 

 the contrast between the clear homogeneous substance of the pseudopodia of the white 

 corpuscle as compared with the granular (reticular?) substance in the body of the 

 cell {vide Fig. 4 and awtea p. 184). 



