The variation 1.e4 c5 2.a4 was invented by the American opening theoretician and author Hugh Edward Myers (1930-2008), inspired by a discussion in 1958 about 1.e4 c5 2.Nf3 Nc6 3.Bb5 with GM Nicolas Rossolimo. The latter liked to combine his pet line with an early a2-a4. Myers' stem games with 2.a4 from the 1960s appeared with instructive comments in Exploring the Chess Openings (1978) and in Myers' autobiography A Chess Explorer (2002). 

My own article http://www.chesscafe.com/text/kaiss65.pdf gives an introduction into the ideas, recommending 2.a4 as part of my semi-unusual repertoire for 1.e4 players. The line remained a rarity, but the Dutch IM Gerard Welling used it with success. In his games Welling follows Myers' main concept: White hopes for a version of the Rossolimo Variation, but in contrast to the beginning 1.e4 c5 2.Nf3 White may be able to choose a different set-up including f4 or develop his king's knight to e2 after, say 1.e4 c5 2.a4 Nc6 3.Bb5 or 2...Bd7 3.Bb5+. Moreover, in contrast to 1.e4 c5 2.Nf3 g6, when White must be familiar with another system, the Myers Variation offers possibilities to steer the game into Rossolimo channels (or d6 Bb5+ channels) even against 2...g6. - To the following game I added some analysis in order to illustrate White's options after 2...g6.