A recent article in Ecography by Zuloaga and Kerr (2016) addressed a prediction of Janzen's (1967) classic contention that mountain passes are higher in the tropics: species assemblages separated by steep thermal gradients are less similar than assemblages separated by small thermal gradients. Their results have some surprising and important additional implications. In the New World, mountain passes are in fact higher in both the tropics and near 55°N. This fact allows a strong test of Simpson's (1964) hypothesis that thermal barriers promote allopatric speciation, which leads to higher species richness. Combined with the results of Zuloaga and Kerr, the Simpson/Janzen hypothesis predicts a peak of richness in southern Alaskan and northern British Columbia. Published data are clearly inconsistent with this prediction. High thermal barriers to dispersal do not necessarily lead to greater species richness.