Las Abejas by Agustín Barrios
Agustín Barrios composed Las Abejas in 1921, the same year he wrote La Catedral. Although the piece is classified as a study and it is said that Barrios rarely performed it for audiences, the song has been recorded and performed by modern artists almost as much as La Catedral. Paquito D'Rivera's arrangement of Las Abejas for his jazz group is a testament to the popular appeal of the song outside the world of classical guitar.
As with La Catedral, you will find many different variations of Las Abejas, making it impossible to pin down a definitive version. Even John Williams took the liberty of changing notes and structure in his recording on the record From the Jungles of Paraguay. In cases where two performances are similar, the fingerings will differ greatly. Instead of imitating other performers, you must explore the work on your own to arrive at a strategy for playing the piece that matches your style. Nevertheless, I believe the fingerings I have provided for the second part of the piece (measures 19–36) are the most natural. Your hand moves through a series of fixed chord shapes around which you are able to play the arpeggios at a rapid speed.
Las Abejas is reputedly an arpeggio study, yet many guitarists play it with a lot of left-hand legato. It is quite easy to play the first measures using left-hand legato, but in order to preserve the spirit of an arpeggio study I have notated measures 3 and 4 as arpeggios that may present some difficulty for the left hand. That is, until you've practiced the measures enough to develop the additional flexibility and strength required for the fingering. Once you master this fingering, which exercises little-used muscles in the left hand, you should find other challenging pieces easier to play. If you opt for left-hand legato, you will not gain such a benefit and may find the rest of Las Abejas, including the seven-fret stretch in measure 37, more time-consuming to learn to execute with ease. That is not to say you shouldn't use left-hand legato in other parts of the piece. I mean only to point out that playing measures 3 and 4 with left-hand legato because it is easier may cause you to miss out on the benefits of playing the measures as arpeggios.
I've tried to make the music easier to read by adding phrasing slurs to the first part and accent marks to the second part. The phrasing slurs are the dotted lines below the staff, where I placed them to avoid cluttering the rest of the notation above the staff. The choice of phrasing is debatable. Its purpose is to avoid getting lost in the stream of eighth notes. The accents in the second part highlight the melody, making it easy to follow the music while playing at breakneck speed. These notes are also more accurately represented as being of longer duration. Despite recommended notation practice frowning upon merging half notes with solid notes, I have merged dotted half notes with eighth notes. It is not a good practice for other instruments, but is idiomatic for the guitar and more readable than showing the notes separately (as one would do when writing for the piano) when they result from the same right-hand stroke.
As is my norm, I've written the music in 12/8 time instead of 4/4 with triplets. I've also included ossias with alternate beginnings and endings. The ossia in measure 7 is meant to connect with the coda ossia in measure 39. Do not go from measure 9 to the coda ossia. Measure 9 connects with the regular measure 39, not the ossia. The ossia from the first two measures shows how John Williams plays the first two measures on From the Jungles of Paraguay. He may have felt it sounded more like a bee. I do not provide fingering for the first two measures because how you play them depends on the sound you prefer.
Changed half-barre in measure 23 to start halfway into the measure instead of the beginning. It's easier to move the hand into position without the barre at the start of the measure.
Placed accent marks below staff, pairing them with second voice to improve legibility.
 I would have liked to have provided chord diagrams for the second part of the piece as I've done for other songs, but it's just too much trouble with LilyPond.
 I'm referring to the arpeggios I notated, which play A and D on the sixth and fifth strings respectively instead of open strings. The arpeggios can be played with open A and D, which I also believe defeats the intent of the study.