The Emperor and Hawaiian volcanic chains are usually assumed to represent a single entity, and to be an exemplary fit to the plume model. However, many features do not conform to this hypothesis. These include: 1) the great “bend” does not result from a change in direction of motion of the Pacific plate, since no such change occurred, 2) southerly motion of ~ 800 km of the melt locus occurred during formation of the Emperor chain, and stopped at the time of the great bend, 3) the chain apparently commenced near a ridge, 4) there is no Hawaiian “plume head”, 5) the melt volume flux varies greatly along the chains and has increased by an order of magnitude during the last 5 Myr, 6) there is no significant heatflow anomaly around the bathymetric swell, 7) there is no measurable thermal rejuvenation near Hawaii, 8) low temperatures relative to those required for a plume are estimated from geothermometry, 9) there is no petrological or geochemical evidence that requires a deep melt source, and 10) seismic evidence indicates relatively normal wave speeds throughout the mantle beneath Hawaii. A large number of seamounts unaffiliated with volcanic chains have similar geochemistry to Hawaii. A fully quantified hypothesis for Hawaii that can explain these features is long overdue.