Articles are meant to be somewhat easy on the general reader, but Hubble tends to get caught in the current of favoritism more than he should, IMO.
“
In the 1920s, astronomer Edwin Hubble made two amazing back-to-back discoveries: Early in the decade, he found that "island universes," now known as galaxies, sit very far away from us; later that decade, he discovered that, on average, all galaxies are receding away from us.”
Yes: Oct. 4, 1923. This was the famous plate image of M31 (Andromeda). Hubble had discovered the first "Cepheid" variable star. He found others and by 1926, IIRC, had over a dozen of what he felt were fairly accurate distance measurements to other “extragalactic nebulae”, as he called them.
But “No”: He didn't "discover" recessional velocities for spiral nebulae. [He, mainly Humason (co-astronomer), produced great works for redshifts, along with Hubble's great work on distances.]
Almost a decade earlier, Slipher (1914) received a standing ovation from the AAS for his discovery of redshift velocities. He held that their great speeds presented these spiral nebulae as “island universes”, as he published several years later. [This term seems to have come from Shapley who referred to globular clusters as “island universes”, partly to counter the cosmic model of Bohlin (1909) who placed globulars in the center of the universe; MW stars in a ring much farther out.]
One of the strange twists of fate seems to be found in Hubble's Cepheid calculations. His 900,000 lyrs. distance for Andromeda was due to his, or anyone's, lack of knowledge for the Pop II Cepheids (i.e. W. Virginis) variables, which are roughly 4x dimmer and far less massive.
But Leavitt's work was of the Cepheids (Pop I), and Hubble seems to have been finding these same Cepheids, so how did he stumble with such a low distance estimate for M31?
One source mentions that he favored Shapley's detailed work in finding "Cepheids" in globular clusters. So, IMO, with his use of the world's greatest telescope at Mt. Wilson, Shapley likely had better P-L data superior to Leavitt, perhaps. But these, proved later by Baade, were the Pop II (W. Virginis) dimmer variables. Hence Hubble thought he was observing distant Cepheids (per Shapley's data) when he was really seeing the brighter Pop I. This would make for an erroneous closer distance calculation for Andromeda. It also produced too fast of a "Hubble Constant", which he seems to have shrunk a little to about 500 kps/Mpc from his initial estimate. [Lemaitre's, first ever expansion constant, was something like 650 kps/Mpc.]
Any comments on this, because this isn't something often mentioned anywhere that I've noticed? There must be something I'm missing in these calculations.