A jet algorithm must specify how to (re-)combine different partons or
towers into a single four-vector. Various recombination schemes have b
een used experimentally to examine the transverse energy profile of je
ts in hadron colliders. Generally, the data is insensitive to which sc
heme is used. However, we argue that the recombination scheme previous
ly used by the D empty set collaboration is expected to have large per
turbative corrections and should not be used for the purposes of makin
g a quantitative comparison with fixed-order perturbation theory.