Angiosperm pollen grains possess walls of remarkable structural complexity, and the architectural forms encountered are often sufficiently specific and consistent to be useful taxonomically. Lindley, von Mohl and others appreciated this systematic potential almost a century and a half ago, and today a comprehensive pollen taxonomy is taking shape with the publication of the
World pollen flora
founded by the late Professor Gunnar Erdtman. Surprisingly, until quite recently this interest in the taxonomic aspects of pollen wall morphology has not been matched by any great concern for the functional significance of even the most conspicuous features, such for example as the deep sculpturing so commonly found in the outer layer of the wall. When Erdtman spent a period in my laboratory in Belfast in the mid-1950s we talked at length about pollen and spore morphology and morphogenesis, but I do not recall that we seriously touched upon the adaptive meaning of the wall architecture. We need not have heen so blind, for there were already many suggestive leads. Wodehouse in his distinguished book of 1935 addressed himself not only to the problems of pollen wall morphogenesis, but noted also relations between structure and dispersal agency. German authors had thought along similar lines, and had already pointed to the functional importance of one type of pollen surface material, that which in some species acts as a binding agent (
Polenkitt
), holding grains together in groups and facilitating attachment to pollen vectors. Much earlier had appeared the work of Green (1894), who seems to have been the first to study the nature of pollen emissions. In an early application of substrate film methods, Green showed that intact moistened pollen grains released various hydrolytic enzymes, and surmised that these played some part in the pollen-stigma interaction, perhaps in germination and pollen-tube growth and nutrition. When Green wrote, Blackley’s famous work on the causes of hay fever was already 20 years old; and by the 1950s it was well established that the active constituents released by pollen were non-dialysable and predominantly protein in nature. Various authors commented on the rapidity with which protein exudates are released from pollen grains, and there was speculation on the sources of such mobile fractions; indeed the facts suggested the possibility that these, too, were held like
Pollenkitt
in sites at or near the surface.
The molecular signatures of compatible and incompatible pollination