This paper attempts to localize the ‘random factor’ at the inception of a spiral phyllotactic pattern, and to discover which other factors may influence leaf initiation at this point.

The study shows that the third leaf of Trigonella foenum-graecum L. can occupy an alternative site, left (L) or right (R) of the first leaf and thereby introduces L-R asymmetry. The mean divergency angle with the second leaf is about 125°. The positions occupied were, within statistical expectations, equal, but in about 4% of seedlings both the L and R site were occupied resulting in ‘symmetries’. L-R asymmetry was established independently of the sense of dorsiventrality.

In the seed only the first leaf was present and the site of the third leaf not yet determined. Removal of one cotyledon when performed during germination (Day 2 and 3 but not on Day 4), resulted in bias away from the remaining cotyledon. Removal of the distal half of a cotyledon did not affect the L-R frequency distribution.

In immature embryos removal of one cotyledon often resulted in a change of the siting of the second leaf, but toward the remaining cotyledon.

In intact seedlings the occupancy of the L or R site was not correlated with differences in weight between the left and right cotyledon, but a rather weak effect of gravity was observed.

The effects of applying some phytohormones to a petiole stump in decotylized embryos were investigated. Morphological variants of primordial development were recorded.

It is concluded that the ‘random factor’ is of internal origin, close to the apex. Auxins which stimulate primordial development, and other inhibitory substances may play a role in leaf siting.