We report the results of spectral and morphological studies of X-ray data from the supernova remnant ( SNR ) W44 .
Spectral analysis of archival data from the Einstein Observatory , ROSAT , and Ginga , covering a total energy range from 0.3 to 8 keV , indicates that the SNR can be described well by a nonequilibrium ionization ( NEI ) model with temperature \sim 0.9 keV and ionization timescale of order 6000 cm ^ { -3 } years .
All elemental abundances are found to be within about a factor of two of their cosmic values , with iron possibly appearing to show significant depletion .
No clear evidence for emission from supernova ejecta can be inferred from the observed metal abundances .
The column density toward the SNR is high – around 10 ^ { 22 } atoms cm ^ { -2 } – as expected given the location of the remnant in the Galactic plane .

In addition to the spectral analysis , we have investigated two different evolutionary scenarios to explain the centrally-brightened X-ray morphology of the remnant : ( 1 ) a model involving the slow thermal evaporation of clouds engulfed by the supernova blast wave as it propagates though a clumpy interstellar medium ( ISM ) ( White & Long 1991 , hereafter WL ) , and ( 2 ) a hydrodynamical simulation of a blast wave propagating through a homogeneous ISM , including the effects of radiative cooling .
Both models can have their respective parameters tuned to reproduce approximately the morphology of the SNR .
The mean temperature of the hot plasma in W44 as determined by our NEI X-ray analysis provides the essential key to discriminate between these scenarios .
Based on the size ( using the well established distance of 3 kpc ) and temperature of W44 , the dynamical evolution predicted by the WL model gives an age for the SNR of merely 6500 yr. We argue that , because this age is inconsistent with the characteristic age ( P / 2 \dot { P } \sim 20000 yr ) of PSR 1853+01 , the radio pulsar believed to be associated with W44 ( Wolszczan , Cordes , & Dewey 1991 ) , this model does not provide the explanation for the center-filled morphology .
We favor the radiative-phase shock model since it can reproduce both the morphology and age of W44 assuming reasonable values for the initial explosion energy in the range ( 0.7 - 0.9 ) \times 10 ^ { 51 } ergs and the ambient ISM density of between 3 and 4 cm ^ { -3 } .