miles which a given craft could search over in an hour, again 

 a simple first assumption was made 5 that this rate of search 

 was equal to the speed of the craft multiplied by twice the 

 range of detection for the gear in question,, The rate of 

 search varies widely with type of craft and from craft to 

 craft , 



As work on the search problem proceeded, it was found 

 that fears concerning the inadequacy of operational reports 

 were well foundedo It had been hoped that empirical curves 

 for the probability of sighting could be computed from the 

 operational data. But It was discovered that many reports 

 did not give both elevation and visibility and the reports 

 which did were not numerous enough to provide a solid basis 

 for computations. There was an even more fundamental diffi- 

 culty. The reports were confined to sightings which" resulted 

 in attacks o There were no corresponding reports sent in on 

 sightings which did not result in attacks . There was good 

 reason to believe that these latter sightings were often the 

 long range sightings and their absence resulted in a one- 

 sided set of facts. 



The research workers, however, went ahead with their 

 problem and gathered such data as they could , From the 

 Eastern Sea Frontier and from the Bay of Biscay area, data 

 were obtained giving the total number of hours flown by the 

 planes and the total number of sightings of all sorts ob- 

 tained. In these areas it was possible to estimate with 

 reasonable accuracy the number of submarines present at a 

 given time, and by also estimating their submergence tactics, 

 one could estimate the average number of submarines which 

 could be sighted by planes in the area. From these data an 

 estimated search rate was obtained for both the Eastern 

 Sea Frontier and the Bay of Biscay area. 



Then came a somewhat dismaying revelation,, The esti- 

 mated effective search rate turned out to be from 1/3 to as 

 little as l/20 of that computed by using the average range 

 of vision times the speed of the plane. The discrepancy was 

 too great to be explained away by ascribing it to inadequate 

 data, or errors of computation. One explanation was Ineffi- 

 ciency in our own search tactics, but part had to be due to 

 the fact that many submarines which should have been seen were 

 not seen. Admittedly rough, as these results of the first 

 search studies were, they po_ ited to the great necessity of 

 obtaining more complete data on sightings, for without them 



60 



