This is very sad. In 1930, both Eva and William James Ellis were in psychiatric hospitals:
"United States Census, 1930," database with images, FamilySearch (https://familysearch.org/ark:/61903/1:1:XC3B-7YD : accessed 26 March 2021), Eva Ellis, Steilacoom, Pierce, Washington, United States; citing enumeration district (ED) ED 87, sheet 9A, line 3, family , NARA microfilm publication T626 (Washington D.C.: National Archives and Records Administration, 2002), roll 2510; FHL microfilm 2,342,244.
The birthplaces (Washington, Denmark, Minnesota) tie this record to others known for Eva.
"United States Census, 1930," database with images, FamilySearch (https://familysearch.org/ark:/61903/1:1:XCQ9-5QD : accessed 26 March 2021), James Ellis, Sedro-Woolley, Skagit, Washington, United States; citing enumeration district (ED) ED 51, sheet 3B, line 71, family , NARA microfilm publication T626 (Washington D.C.: National Archives and Records Administration, 2002), roll 2513; FHL microfilm 2,342,247.
Again, the birthplaces (Kansas, Missouri, Missouri) match what we know for William.
Here's the death certificate matching the FindAGrave memorial that might be William's:
"Washington Death Certificates, 1907-1960," database, FamilySearch (https://familysearch.org/ark:/61903/1:1:N356-N9T : 10 March 2018), W. James Ellis, 19 Jun 1932; citing Tieton, Yakima, Washington, reference 222, Bureau of Vital Statistics, Olympia; FHL microfilm 2,022,969.
Other than the occupation (laborer), there's not much to match to Eva's husband. He died of a stroke, not his first, and had not worked for several years.