INTRODUCTION 



these variations are not usually of great importance. Relative 

 sizes may be estimated, and as long as we are dealing with struc- 

 tures of unknown molecular composition, subjected to unknown 

 alterations during preparation, the significance of differences in 

 minute dimensions is remote. Only for a very few types of 

 materials (striated muscle, collagen, myelin) are we in a position 

 to care deeply whether a given interval is measured at 5 m^u, or at 

 twice that. To avoid any implication that figures are accurate to 

 the last Angstrom unit, all measurements will be given here in 

 microns and millimicrons. 



