332 



SCIENCE 



[N. S. Vol. L. No. 1292 



Later Lowell* tried out a similar experiment 

 with the result of finding his wire visible 

 easily at 0".89, with some difficulty at 0".83 

 and glimpsed down to 0".69. Evidently his 

 contrast conditions were less good than Bar- 

 nard's. A further test by Slipher and Lamp- 

 land^ showed the wire disappearing from cer- 

 tain vision at 0".86, while a dark blue line on 

 a white disk held down to 0".83. "W. H. 

 Pickering'^ experimenting with a dark human 

 hair, against open sky found it easily visible 

 when subtending an angle of 1".13, easily 

 glimpsed at 0".9Y, occasionally glimpsed at 

 0".83, and quite invisible at 0".Y2. 



Taking up the converse case the writer first 

 tried a German silver wire 0.01 inch diameter 

 stretched zigzag in lengths of several feet over 

 a dark plank bulkhead. The reflectivity of 

 this varied from about 0.06 to 0.12, i. e., a 

 very dark gray. The test was in fidl sun- 

 shine and the observers, the place being the 

 range of the Massachusetts Rifle Association, 

 were a group of riflemen, keen of sight and 

 experienced in close observation. The terrain 

 was laid off in 50-foot spaces and the results 

 were as follows: Wire vanished across lighter 

 parts of background at 75 feet (2".3) while 

 across the darkest of that background the wire 

 persisted up to 200-250 feet, beyond which it 

 was invisible save for specular glints especially 

 at twists. To summarize: 



Angular 

 Diameter Appearance 



l".ll. .Parts against dark baekground were plain. 

 0".86. .Portions seem distinctly but not steadily. 

 0".69 . . Visible at specular spots, difficultly glimpsed 



elsewhere. 

 0".46. .Visible by specular reflection only. 



A second test with some of the same ob- 

 servers was made, using a background of black 

 paper (coefficient .045), white thread 0.008 

 inch in diameter, and drawn tungsten wire 

 0.005 diameter. The paper was nailed to the 

 former bulkhead and wire and thread stretched 

 zigzag as before. Observed in bright sky- 

 light, also in moderately bright sunshine. 

 The wire was vicible with difficulty to one 



, ^ Bulletin Lowell Ohs., No. 2. 

 , Lowell Ols. Bull. No. 10. 

 6 Pop. Ast., 23, 578. 



observer at 150 feet and beyond this disap- 

 peared utterly. To the others it could be 

 fairly made out only at 100 feet. The thread, 

 which was much brighter than the wire, began 

 to be lost in parts at 200 feet, but in sun- 

 shine held rather indistinctly but unmis- 

 takably to 300 feet. When the sun went in 

 the thread was lost at about 200 feet. To 

 simamarize again 



Angular Diameter Appearance 



Wire 0".86 Limit for all but one observer. 



0".57 Fairly seen by one observer. 



Thread 0".92 Network distinct all over. 



, 0".69 Parts distinct in sunshine. 



0".55 Parts evident in sunshine. 



I 0".46 Near limit of visibility. Only 



small bits of network seen. 



In the case of the thread the brightness 

 contrast between thread and background was 

 about 16 : 1. With brilliant sunshine and a 

 background of even deader black there might 

 have been a slight further gain, but we were 

 evidently close to the limit. It is rather note- 

 worthy that there should be so near an agree- 

 ment throughout as between dark on bright 

 and bright on dark, but barring specular direct 

 reflection the brightness contrast which deter- 

 mines visibility is not widely different in the 

 two cases, and the minimum visibile for a 

 linear object with strongly contrasted back- 

 ground would appear to be about 0".5 =i=. It 

 is certainly less than 1/50 the minimum 

 visihile for a round spot giving similar con- 

 trast, a remarkable evidence of the efficient 

 coordination of retinal impressions. 



Boston Louis Bell 



SCIENCE 



A Weekly Journal devoted to the AdvEincement o£ 

 Science, publishing the oflficial notices emd pro- 

 ceedings of the American Association for 

 the Advancement of Science 



Published every Friday by 



THE SCIENCE PRESS 



LANCASTER. PA. GARRISCSsI. N. Y. 



^4EW YO^C, N. Y. 



Eoteted in the post-o^e at Limcastcr, Pa., u Mcond dan mtttwi 



