I've just learned about an indigenous Brazilian flute which, at first glance, looks almost exactly like an indigenous Southern African flute (not the part of Africa where slaves were taken to the Americas, so they must have been independent inventions !). I can't find any recordings or video of either flute online, but I think they're only visually/superficially similar, which is fascinating
this is a traditional flute of the Apinajé, a branch of the Gê-speaking indigenous peoples from central Brazil (link to another example at the Swedish Ethnographic museum here). it is (in German) described as : "Pfeife mit kleinem Schallkürbis der Träger des Namens Tamgaága" (the Museum lists "Tamgaága" as the local name of the flute, but I think that was actually the name of the original owner. the word has a few hits on google in Portuguese articles and seems to be a personal name, and it has 0 hits for anything flute-like. anyone who can read Portuguese please verify)
and these are the "ombgwe" and "khumbgwe", traditional flutes of the Shona-Karanga and Venḓa peoples of Zimbabwe and South Africa (from P. R. Kirby, 1934 (2nd ed. 1968) The Musical Instruments of the Native Races of South Africa, Plate 45)
Although they look alike superficially, I think they're actually played from the opposite ends. below is the archival card from the Göteborg Museum for the Apinajé flute :
it suggests there are some more details in "comment no. 5 in the original catalogue," which I'm pretty sure is this one from 1931 (link to readable version on Google Books here). although I can figure out a brief catalogue entry, I don't actually speak Swedish so finding the relevant passage in that whole book has proved difficult...
anyway. the point is that looking at the sketch and hole placement on the calabash resonator, I'm fairly sure the flute is played by blowing across the top of the reed pipe, just as you would play panpipe for example, and the pitch is changed using the finger holes on the calabash at the bottom. if the finger holes were at the top, it would make the remaining pipe length irrelevant so I can't imagine it's played the other way around. also think of the western oboe family of instruments, which have a bulbous bell at the bottom to affect the timbre.
now, consider the Shona-Karanga ombgwe. Kirby provides a photo of one being played in his book cited above, and Margaret de Lange describes the playing technique in the catalogue of the old Africana Museum in Johannesburg.
the ombgwe is played with the gourd (or nsala fruit shell) at the top, blowing across a mouth hole just as you would on a Chinese xun for example, and the reed pipe extends downwards. the finger holes are once again on the bottom of the instrument, ie on the reed tube of the ombgwe