The heating and cooling of the interstellar medium ( ISM ) allow the gas in the ISM to coexist at very different temperatures in thermal pressure equilibrium .
The rate at which the gas cools or heats is therefore a fundamental ingredient for any theory of the ISM .
The heating can not be directly determined , but the cooling can be inferred from observations of C ii* , which is an important coolant in different environments .
The amount of cooling can be measured through either the intensity of the 157.7 \micron [ C ii ] emission line or the C ii* absorption lines at 1037.018 Å and 1335.708 Å , observable with the Far Ultraviolet Spectroscopic Explorer and the Space Telescope Imaging Spectrograph onboard of the Hubble Space Telescope , respectively .
We present the results of a survey of these far-UV absorption lines in 43 objects situated at \mid b \mid \gtrsim 30 \arcdeg .
Measured column densities of C ii* , S ii , P ii , and Fe ii are combined with H i 21-cm emission measurements to derive the cooling rates ( per H atom using H i and per nucleon using S ii ) , and to analyze the ionization structure , the depletion , and metallicity content of the low- , intermediate- , and high-velocity clouds ( LVCs , IVCs , and HVCs ) along the different sightlines .
Based on the depletion and the ionization structure , the LVCs , IVCs , and HVCs consist mostly of warm neutral and ionized clouds .

For the LVCs , the mean cooling rate in erg s ^ { -1 } per H atom is -25.70 ^ { +0.19 } _ { -0.36 } dex ( 1 \sigma dispersion ) .
With a smaller sample and a bias toward high H i column density , the cooling rate per nucleon is similar .
The corresponding total Galactic C ii luminosity in the 157.7 \micron emission line is L \sim 2.6 \times 10 ^ { 7 } L _ { \odot } .
Combining N ( C ii* ) with the intensity of H \alpha emission , we derive that \sim 50 % of the C ii* radiative cooling comes from the warm ionized medium ( WIM ) .
The large dispersion in the cooling rates is certainly due to a combination of differences in the ionization fraction , in the dust-to-gas fraction , and physical conditions between sightlines .
For the IVC IV Arch at z \sim 1 kpc we find that on average the cooling is a factor 2 lower than in the LVCs that probe gas at lower z .
For an HVC ( Complex C , at z > 6 kpc ) we find the much lower rate of -26.99 ^ { +0.21 } _ { -0.53 } dex , similar to the rates observed in a sample of damped Ly \alpha absorber systems ( DLAs ) .
The fact that in the Milky Way a substantial fraction of the C ii cooling comes from the WIM implies that this is probably also true in the DLAs .

We also derive the electron density , assuming a typical temperature of the warm gas of 6000 K : For the LVCs , \langle n _ { e } \rangle = 0.08 \pm 0.04 cm ^ { -3 } and for the IV Arch , \langle n _ { e } \rangle = 0.03 \pm 0.01 cm ^ { -3 } ( 1 \sigma dispersion ) .

Finally , we measured the column densities N ( S ii ) and N ( P ii ) in many sightlines , and confirm that sulphur appears undepleted in the ISM .
Phosphorus becomes progressively more deficient when \log N ( H i ) > 19.7 dex , which can either mean that P becomes more depleted into dust as more neutral gas is present , or that P is always depleted by about -0.3 dex , but the higher value of P ii at lower H i column density indicates the need for an ionization correction .