ULTRA-SHORT-WAVE PROPAGATION 265 



more or less unobstructed optical path existed along Columbus 

 Avenue. 



While a detailed analysis of the attenuation of ultra-short waves 

 over paths as complicated as those considered in this paper is im- 

 possible, the known facts indicate several general characteristics that 

 seem worthy of mention for further experimental investigation. The 

 fact that the field was found to be approximately proportional to the 

 height of the antenna above the roof level of the surrounding buildings 

 (in Fig. 4), rather than to the height above the ground, as in the case 

 of transmission over level terrain free from buildings, confirms the 

 expectation that the "ground" conditions in the immediate vicinity 

 of the fixed terminal would play an important part in determining the 

 magnitude of the received field strength. 



It is perhaps correct to assume that for the fixed terminal the height 

 to be substituted in equation (5) should be the height above the roof 

 rather than the height above ground. This would reduce the "level 

 terrain" curve of Fig. 2 by 10 db. 



Figures 6, 7 and 9 show marked wave interference patterns which 

 indicate reflections from a multiplicity of points in the immediate 

 vicinity of the mobile terminal. Besides these variations in the magni- 

 tudes of the fields observed at points in close proximity to each other, 

 which are undoubtedly caused by reflections from irregularities in the 

 immediate vicinity of the terminal, there are the variations represented 

 by the spread of the points about the mean curve (Fig. 2). That these 

 variations may be attributable to conditions local to the terminal is 

 indicated by the fact that the increase in field on the far side of salt 

 water and the decrease in the field on the far side of hills, etc., do not 

 persist at further distances. Even if the irregularities of the contours 

 of Fig. 8 were removed the contours would not be concentric circles 

 about the transmitter. At this stage in the development it would be 

 unwise to attempt to say how much of the deviation of the contours 

 from circles may be attributable to directional characteristics at the 

 fixed terminal and how much may be due to the intervening terrain. 

 Statistically speaking, however, it is safe to say that the additional 

 attenuation attributable to deviations from level terrain such as is 

 produced by the presence of buildings, is independent of the length 

 of the transmission path, so that there is no absorption in the usual 

 meaning of the word. Another experimental result that points to 

 the possibility that the effect of the buildings is local may be deduced 

 from the data presented by Jones ^'^ and by Trevor and Carter.'" 

 The latter showed that the more distant points lie on the inverse- 

 square-of-distance curve expected for transmission over level terrain 

 free from buildings. The nearby points, however, lie below this 

 curve, indicating that the major efi'ect of the buildings is a local one. 



