A microheater was used to produce a temperature gradient within the mitotic spindle of living cells. The slope of the temperature gradient was estimated from thermal conductivity calculations and confirmed by measurements of spindle birefringence and by experiments on striated muscle. When the microheater was placed at one spindle pole or at one group of kinetochores, the gradient was steep enough to cause a large difference in birefringence between the two half-spindles, but the velocity of chromosome movement in anaphase was nearly the same in the warmer and cooler half-spindles. When the heater was shifted from the pole toward the interzone, the average velocity of chromosome movement increased approximately two-fold but was, again, nearly uniform in the two half-spindles. The rate of spindle elongation was especially sensitive to the site of heating, increasing ten-fold when the heater was shifted from the pole to the interzone. Regardless of heater position, the rate of chromosome movement was determined largely by the temperature of the coolest spindle region -chromosomes in the warmer half-spindle moved more slowly than expected from estimates of the temperature in that region. Since the microheater produces a substantial temperature gradient within the spindle, the near uniformity of chromosome velocity in both half-spindles must be due to some biological property of the spindle. Two very different explanations for the results are considered the most likely. According to one explanation, the near uniformity of velocity in both half-spindles is determined by the structure of the interpolar spindle, while changes in velocity involve force producers located both in the half-spindles and in the interzone. On the other explanation, the velocity is nearly the same in both half-spindles because the force producers are located exclusively in the interzone (Margolis et al., 1978). © 1979 Springer-Verlag.
