Detailed computations reveal a large difference of hinetic temperature T between H i and H ii regions. In observed H II regions T probably exceeds 7500 K, a lower limit established by observations of [0 ii] X 3727, and is probably not greater than 13,000 K. The theoretical give actual temperatures ranging between 5000 and 10,000 near B stars and between 7000 and 13,000 near 0 stars; the exact value depends on the ratio of C, N, and 0 ions to protons and on the cross-section for electron excitation of these ions. In H i regions the sources of energy gain are relatively weak; the temperatures of clouds are reduced to about 60 K by three processes: electron excitation of the low-lying levels in C II and Si ii, inelastic collisions between H atoms and grains, and inelastic collisions between H atoms and H2 molecules. Even under somewhat extreme assumptions, the temperature in an H i cloud does not exceed . Values of the time tE required to approach the equilibrium temperature are very short if the density is high. If the density of H atoms or ions is 10' per cubic centimeter, t5 is only about 10' years in both H i and H ii regions. However, if the density of H atoms or ions is taken to be 10 , a value which may represent rarefied regions between clouds, the values of t5 are about 106 years in H ii regions and 10 years in H i regions. As a result of the low temperatures found in H i clouds, the aver age charge on a dielectric interstellar grain is very small but may rise to several volts in H ii regions.