1890.] The Modern Weather Bureau. 27 



In this latter room the lithographer has already prepared a supply of 

 the blank base map. He has also a solid metallic form, of the size of this 

 map, in which are cut slots for the insertion of square type or dies. 

 He puts in each slot the type that represents the wind and weather as 

 you see them here printed, and prints one copy of these symbols in 

 " transfer ink." He then transfers this " symbol-print " to a large 

 lithographic stone and also transfers to the same stone the isobars the 

 isotherms, and the left and the right hand corner tables. All this work 

 is being carried on simultaneously by several and he is ready to print 

 the special morning map all that you see here in red on the green base 

 map, in about 20 minutes after he receives the data from the prediction 

 room. The schedule of time was formerly about as follows for these 

 7 a.m. maps : observation at 7 a.m., telegraphy 8 to 8.30, prediction 

 finished at 9.30, map printing begun at 10 a.m. 



For the 8 a.m. observation now employed the time is shortened 

 and the printing begins about 9.30. You will understand that 

 this same routine is again gone through from 4 to 7 p.m., and again 

 from 10 p.m. to 1 a.m.,, and it has so gone on without a single 

 intermission for holidays or Sundays, or a single failure since January 

 1871, so far as maps and predictions are concerned, and since June 1871 

 so far as printed maps are concerned. 



In answer to a question from Dr. Gill I would say that the clerks 

 rotate in their duties, but that the same predicter must be on hand for 

 at least a month, when he may be relieved by another. It has been 

 found quite impracticable to have different men do the morn- 

 ing and night predictions, one person must take the entire responsibility 

 lor the time being. He lives as it were in continuous contemplation 

 of the weather ; he comes to the study of each nevf weather map 

 with a vivid remembrance of the conditions shown on the preceding 

 map ; the eight or twelve hour intervals do not allow him to forget any- 

 thing, as might be the case with twenty -four hour intervals. He 

 comes to the new maps expecting to tind certain changes, and if any of 

 these have not occurred then these are the features that demand especial 

 attention and study. The Signal predicting officers were formerly 

 never known to tire public by name, but of late years their names have 

 been published in the Monthly Weather Review in connection with a 

 statement of the percentages of verification. s I take pleasure in adding 

 that Captain Robert Craig and Captain H. H. Dunwoody, who in 1872 

 began to alternate with me as predicters, are still engaged in the Signal 

 Office, in that or cognate work. Captain Dunwoody is regarded as our 

 best predicter at present. 



But the printing of our weather maps is merely the official record of 

 our work and does not suffice to make our predictions practically useful 

 to the people ; it is this latter result that is a most vital feature in the 



