Modifications are shown of Swanson's original instrument for determining heat pulse velocity in woody plants. These modifications stabilize the instrument from −50° to +80°C and allow simultaneous multipoint sampling. Laboratory experiments are reported which demonstrate that heat pulse velocity is closely related to volume of water forced through stem sections. Field experiments with juniper trees show that (1) on clear days, the heat pulse velocity increases rapidly from sunrise to midmorning and remains high until sunset; low rates prevail at night; (2) leaf wetting produces an immediate decrease in heat pulse velocity; rates remain low while leaves are wet, then increase rapidly when leaves dry; (3) within 24 hours after dry soil is irrigated, heat pulse velocity increases two to five times; and (4) natural soil drying, even with high air temperatures, causes a decrease in heat pulse velocity. These results tend to verify what has been heretofore mainly conjecture based on the measurement of external variables surrounding a tree.