He with a half-
life of 12.43 years [ Taylor and Roether, 1982]. Analysis for
tritium is done in the laboratory via gas proportional counting
[cf. Östlund and Dorsey, 1977] and mass spectrometric
measurement [cf. Clarke et al., 1976]; the latter method is
also used for 
He [cf. Lott and Jenkins, 1984]. Tritium is
delivered to the oceans via vapor exchange, rain, and continental
runoff. It is most commonly found in the oceans in the form HTO
(proton-tritium-oxygen) [ Östlund and Mason, 1977]. Pre-
bomb naturally occurring background levels of tritium are estimated
to be about 0.2 to 0.5 TU in surface waters [ Begemann and
Libby, 1957; Dreisigacker and Roether, 1978], where 1 TU =
10
*[T/H]. The bomb tests of the 1950s and particularly the
early 1960s increased the atmospheric tritium inventories by at
least two orders of magnitude [ Craig and Lal, 1961;
Eriksson, 1965]. Most of the bomb tritium entered the surface
waters of the high northern latitude oceans within a few years
causing a tritium maximum in surface waters in about 1964. Several
studies have used models to simulate the tritium delivery to the
North Atlantic Ocean as a function of time [e.g. Dreisigacker
and Roether, 1978; Weiss et al., 1979; Weiss and
Roether, 1980; Koster et al., 1989]. Recently Doney et
al. [1993] formulated a model for the period 1950-86 that agrees
to within 
10% with the water column inventories. The model
results show that starting in the mid-1970s, the annual tritium
input from the Arctic exceeded that of all the other sources
combined (Figure 2a).