The gas-to-dust mass ratios found for interstellar dust within the Solar System , versus values determined astronomically for the cloud around the Solar System , suggest that large and small interstellar grains have separate histories , and that large interstellar grains preferentially detected by spacecraft are not formed exclusively by mass exchange with nearby interstellar gas .
Observations by the Ulysses and Galileo satellites of the mass spectrum and flux rate of interstellar dust within the heliosphere are combined with information about the density , composition , and relative flow speed and direction of interstellar gas in the cloud surrounding the solar system to derive an in situ value for the gas-to-dust mass ratio , R _ { g / d } =94 ^ { +46 } _ { -38 } .
This ratio is dominated by the larger near-micron sized grains .
Including an estimate for the mass of smaller grains , which do not penetrate the heliosphere due to charged grain interactions with heliosheath and solar wind plasmas , and including estimates for the mass of the larger population of interstellar micrometeorites , the total gas-to-dust mass ratio in the cloud surrounding the Solar System is half this value .
Based on in situ data , interstellar dust grains in the of 10 ^ { -14 } to 10 ^ { -13 } g mass range are underabundant in the Solar System , compared to an MRN mass distribution scaled to the local interstellar gas density , because such small grains do not penetrate the heliosphere .
The gas-to-dust mass ratios are also derived by combining spectroscopic observations of the gas-phase abundances in the nearest interstellar clouds .
Measurements of interstellar absorption lines formed in the cloud around the solar system , as seen in the direction of \epsilon CMa , give R _ { g / d } =427 ^ { +72 } _ { -207 } for assumed solar reference abundances , and R _ { g / d } =551 ^ { +61 } _ { -251 } for assumed B-star reference abundances .
These values exceed the in situ value , suggesting either grain mixing or grain histories are not correctly understood , or that sweptup stardust is present .
Such high values for diffuse interstellar clouds are strongly supported by diffuse cloud data seen towards \lambda Sco and 23 Ori , provided B-star reference abundances apply .
If solar reference abundances prevail , however , the surrounding cloud is seen to have greater than normal dust destruction compared to higher column density diffuse clouds .
The cloud surrounding the Solar System exhibits enhanced gas-phase abundances of refractory elements such as Fe ^ { + } and Mg ^ { + } , indicating the destruction of dust grains by shock fronts .
The good correlation locally between Fe ^ { + } and Mg ^ { + } indicates that the gas-phase abundances of these elements are dominated by grain destruction , while the poor correlation between Fe ^ { + } and H ^ { \circ } indicates either variable gas ionization or the decoupling of neutral gas and dust over parsec scalelengths .
These abundances , combined with grain destruction models , indicate that the nearest interstellar material has been shocked with shocks of velocity \sim 150 km s ^ { -1 } .
If solar reference abundances are correct , the low R _ { g / d } value towards \lambda Sco may indicate that at least one cloud component in this direction contains dust grains which have retained their silicate mantles , and are responsible for the polarization of the light from nearby stars seen in this general region .
Weak frictional coupling between gas and dust in nearby low density gas permit inhomogeneities to be present .