12 



Greene and delivered to the Patrol vessel within a few hours after the 

 last station is occupied. This phase of Ice Patrol activity has been 

 in operation since 1932, and the resulting current charts, about one 

 a month, have been used with such success by the Ice Patrol that 

 their accuracy and reliabiUty, within the limits that experience has 

 dictated, are somewhat taken for granted by the officers of the Patrol. 

 Therefore, it would seem a healthy thing, now and again, to actually 

 put down what results are obtained and to illustrate the stated relia- 

 bility by plotting some observed berg drifts against the stream lines 

 of contemporary current maps. Such illustrations are shown in fig- 

 ures 2, 3, and 4. The last time a discussion of this subject was 

 taken up was in 1935. (See Bulletin 25, pp. 8 to 10.) It was stated 

 there that any single oceanograpliic station was good for about 15 

 days; and that as it now requires about 10 days to complete one map, 

 the first portion of the chart completed will be depended on only 

 about 5 days after completion of the whole network. Tliis means 

 that the reliability of the map, assuming that it was begun in the 

 north, will fade progressively from north to south beginning 5 days 

 after the last station was taken and ending approximately 15 days 

 after, at the southern section. 



Bearing these facts in mind, the reader is referred to figure 2. To 

 begin with, the berg whose drift is shown was, when first reported 

 April 10, outside the limits of current map (fig. 45); as the most 

 northern section crossed the cold current in about latitude 44°40' N. 

 So that although the correspondence, in the matter of direction, at 

 least, seems excellent from the 10th to 12th, it is only coincidence; 

 the current strength and direction outside map limits being a matter 

 of conjecture. The berg's drift from April 12 to 14 was slightly 

 across the isobaths, at first, to the south southeast and over the 2 

 days it averaged 20 miles per day instead of 14.5 miles per day as 

 the chart shows. Both these effects could be explained as the results 

 of the strong gales which blew from the northwest on April 7, 8, and 

 9. (See fig. 19.) The wind components accelerating the current are 

 seen to be much more lasting and more pronounced than the devia- 

 tion in direction, as illustrated by the drift from April 14 to 18. Dur- 

 ing these 3 days the berg followed the stream lines fairly well, though 

 the rate of drift was still excessive, the average being a Httle over 

 1 knot instead of 0.3 knot. From the 18th no fault can be found 

 with the direction of drift, but the velocity is still in excess, though 

 decreasing, until the 20th Remembering, however, the facts in the 

 paragraph above, it is realized that the current map has already ex- 

 hausted its period of reliabiUty. For this reason no stream lines are 

 indicated in figure 2 west of longitude 51° W. A comparison of the 

 berg's drift from April 20 until it melted April 29 with figures 45 and 

 46 will show that it drifted in a pattern which appears to be an inter- 



I 



