Subject: Why would stranded wire cause more signal attenuation than solid?

Electron flow is reported to occur primarily on the outer surface of 
a conductor, and yet stranded wire with more wire surface area is 
sold with warnings to keep the lenght short, <20feet, due to high 
signal attenuation. I see this warning specifically for the new CAT5 
level 6 wiring being employed for 100mb 350Mhz computer signals. 
Please explain how the physical benefit of greater surface for 
conducting, provided by stranded wire, becomes a 10x handicap when 
frequency increases. I believe this will be related to the lower 
capacitance of solid wire at 14pF/foot for 24ga. How is the stranded 
wire capacitance calculated, and how does signal frequency fit into 
the attenuation equation. It also appears that the manner of 
stranding could be "bunched" or "compressed" or "compacted"; and the 
number of strands, and size of strands within a wire, and the number 
of twists per foot could all relate to the attenuation. "University 
Engineering" P661-665, explains two wire strands and "characteristic 
impedance" based on self-inductance and capacitance, but not signal 
frequency. Can anyone provide a more comprehensive equation for 
"characteristic impedance"? I can win a nickel! 

Re: Why would stranded wire cause more signal attenuation than solid?