Now looking at the diagrams on Page 4. The right figure (female BNC) shows reference plane is 0.327-0.335 inches "back" from the outer edge of the female connector.
I'm trying to understand, so please bear with (and correct) me --- I gather that at ~3 pSec/mm, the physical distance between the BNC short "short" position (where the center conductor meets the shield) and the BNC female reference plane position would be based on a one-way delay of 113.5/2 = 56.75 pSec, then the distance would be 56.75/3 = 18.92mm (about 0.75 inch). Looking at the Page 4 drawing female connector dimensions, the distance from the reference plane to the outer edge of the female connector is given as 0.327-0.335 inch, so 0.75 inch is a bit over double (that is, way out beyond the outer edge of the female connector.) Yet when I look at a BNC male short, there is very little metal beyond the outer edge of the female connector until the back edge of the BNC male short's body, so I can't visualize how the 0.75 inch one-way electrical distance fits with this.
In any event, why is delay given as a negative number? If I were to consider the male offset short from HP 85033D calibration kit, the (one-way) offset delay for that component is specified by HP as about 31.8 or 31.9 pSec (I don't have the exact number in front of me - Dr. Kirkby I'm sure would know). Why is that delay specified as a positive value (i.e., what one would enter into an HP 8753 that had no existing standards data for the component?)
Thanks for your patience and corrective efforts!