Henny Cathey and I were running some low kV FE epma experiences recently at ASU (we'll be presenting this at the December 2014 AGU meeting in San Francisco). In preparation to run sub-micron domains on tiny "sieve amphiboles" [actually may not be amphiboles] in pseudotachylite, we spent over a day calibrating on various standards. Particularly the Kakanui hornblende as we assumed our target phase would be amphibole. One of the arguments for getting a field emission probe is that you can now see so much better what you are probing. Boy, is this the truth. The Kakanui is chock-full of micron to sub-micron 2 or 3 phase inclusions: apparently an Fe-oxide, a silica-rich phase (glass I am guessing) and possibly a vapor bubble (or the hole left by it). See attached image. Which led to a literature search describing its field occurrence...presumbly these are remanent melt inclusions of this volcanic phase, which also base evidence of re-heating. So, yes, it is no surprise that it exhibits some degree of heterogeneity.