We have combined measurements of the kinematics , morphology , and oxygen abundance of the ionized gas in I Zw 18 , one of the most metal-poor galaxies known , to examine the star formation history and chemical mixing processes . Deep H \alpha imagery shows diffuse emission and a partial shell extending well beyond the main two knots of continuum emission . We have explored the kinematics of this ionized gas using longslit , echelle spectroscopy of the H \alpha line . We find the unambiguous signature of a supergiant shell southwest of the galaxy and weak evidence for a second bubble northeast of the galaxy . The axial symmetry of these shells and the asymmetry in their H \alpha line profiles suggest that they comprise the lobes of a single bipolar bubble expanding at \sim 30 - 60 km s ^ { -1 } . Higher velocity gas is found near the small shell immediately west of the northwest HII region . Although an unresolved X-ray source is discovered near the northwest HII region in archival ROSAT PSPC data , we argue that hot interstellar gas associated with the superbubbles does not produce all the X-ray emission . Oxygen emission lines are detected up to \sim 1 kpc from the NW HII region along the bubble ’ s polar axis , so this diffuse , ionized gas has been polluted with gas processed by stars . Measurements of the O/H abundance ratio in the inner nebula show surprisingly little variation considering the apparent youth of the galaxy . We describe the dynamical evolution of the superbubble using a simple wind-blown bubble model . To test the hypothesis that the dynamical age of the bubble measures the duration of the starburst in I Zw 18 , we compute the photometric properties of a starburst with the same age as the superbubble . We find that star formation commencing 15 - 27 Myr ago and continuing at a rate of 0.017 – 0.021 { ~ { } M _ { \odot } } ( of 1 – 100 { ~ { } M _ { \odot } } stars ) per year can both power the gas dynamics and produce a fair match to the integrated optical properties of I Zw 18 . The total mechanical energy returned to the interstellar medium by stellar winds and supernovae , 7 – 30 \times 10 ^ { 53 } ergs , is insufficient to eject the entire interstellar medium . However , the corresponding mechanical energy injection rate is high enough to drive the superbubble shell out of the HI gas cloud , and “ blowout ” will allow the hot ISM to escape in a galactic wind . This supports the idea that metal-enriched winds play a prominent role in the chemical evolution of dwarf galaxies .