Draw a cooling coil's process on the chart and it runs as a straight line from the entering condition down toward the saturation curve. The point where that line reaches the curve is the apparatus dew point, the ADP: the effective temperature of the wet coil surface. The air never quite gets there, and the model that explains why is the simplest one in the chapter.

Treat the coil as a second mixing box. One share of the air touches the surface and leaves saturated at the ADP. The rest slips between the fins untouched. That untouched share is the bypass factor: $\mathrm{BF} = \dfrac{t_{la} - t_{adp}}{t_{ea} - t_{adp}}$, read aloud B-F equals t-l-a minus t-a-d-p, over t-e-a minus t-a-d-p. $t_{ea}$ is the entering-air dry bulb, $t_{la}$ the leaving-air dry bulb and $t_{adp}$ the apparatus dew point, all in °C. $\mathrm{BF}$ is a bare fraction between 0 and 1. Fix the subscripts once: ea is entering air, la is leaving air, adp is the coil's own cold end. Both differences are measured FROM the ADP.

Read the other ways, the same line answers the design questions. The leaving air is $t_{la} = t_{adp} + \mathrm{BF}\,(t_{ea} - t_{adp})$: start at the ADP and climb back by the bypass share of the gap. It is the mixed-air relation again, with $\mathrm{BF}$ as the weight on the warm stream. The share that did reach the surface is the contact factor, $1 - \mathrm{BF}$. The two always add to one, and quoting one for the other is the classic slip.

Two nuggets. The bypass factor belongs to the coil AND its airflow: more rows and closer fins lower it, and a faster fan raises it, because each parcel of air spends less time against the metal. That is why a coil that dried the air properly at design flow stops doing so when someone speeds the fan up. And the ADP is not the chilled-water temperature. It is an equivalent surface temperature read off the chart, and it always sits a few degrees above the water inside the tubes.