188 



PROCEEDINGS OP THE AMERICAN ACADEMY 



INDIVIDUAL EKKOES, 



It is somewhat doubtful whether the mean values given, represent 

 the actual errors of the spaces. There are some indications that a 

 shrinkage of the film of varnish has occurred since it was first applied, 

 and this action may have produced some effect upon the graphite with 

 which the lines are filled. Although graphite is an impalpable pow- 

 der, I have seen many instances in which it has been lifted in mass 

 from the filled lines and thrown a distance as great as one thousandth 

 of an inch without breaking the continuity of the particles. This 

 action seems to take place only when the lines and the filling are pro- 

 tected by a thin cover-glass, closely cemented to the slip on which the 

 lines are ruled. Sometimes an explosion seems to take place, scat- 

 tering the graphite in all directions, leaving it in curves having nearly 

 a uniform shape. I have never been fortunate in seeing this action, 

 but in the case of one ruled plate an actual observation limits the time 

 within which the explosion must have occurred to about ten days. 

 In this case the lines remained perfect for about four months after 



