Just as Morris seemed to have established the value of agricultural research, he left his post. In mid-1879, the Colonial Office appointed him as the new director of the botanical garden in Jamaica, leaving the planters once again without the support of a scientist. At first, the island’s governor, Sir James Robert Longden, balked at hiring a replacement. He cited Morris’s success as a reason for not appointing a replacement. Morris, argued Longden, “had exhausted the history of the Hemileia.” The planters, he continued, “knew what they had to do and the mode of carrying it out.” That work “belonged to the practical planters rather than scientific men.”55
Even as some planters celebrated Morris’s achievements, others voiced caution. Morris himself had argued that the results of his experiments could only be confirmed after a full growing season. He had left for Jamaica before this, and over the remainder of the season, it became apparent that the chemical treatments Morris had recommended did not, in fact, control the rust. This did not destroy the planters’ newfound faith in science, though. Ceylon’s chamber of commerce requested that the home government appoint “another gentleman of possible equal qualifications and attainments to Mr. Morris.” The Colonial Office once again asked Thiselton-Dyer at Kew to recommend a suitable candidate. He recommended another young scientist named Harry Marshall Ward, who had studied natural science at Cambridge (where he graduated with a first-class degree in 1879) and Würzburg, where he had studied under Anton de Bary and Julius von Sachs, two leading proponents of the new botany. When he returned to the UK, he worked at the Jodrell Laboratory at the newly founded center for experimental plant biology at Kew.56 Thiselton-Dyer and Hooker recommended that Ward be sent to Ceylon on a two-year contract. This was enough time, they felt, for Ward to study coffee over several growing seasons. That would allow him to establish where Morris had gone wrong and—they hoped—to find an effective cure.
Ward brought the new botany to bear on solving the problems of the coffee rust. Over 1880 and 1881, he conducted a wide range of systematic and comparative observations and experiments aimed at understanding the fungus’s life cycle, its epidemiology, its impact on the coffee tree, and potential control measures. In these two years, he produced three important reports for the government of Ceylon detailing the experiments and his findings. He also produced two scientific papers for the Quarterly Journal of Microscopical Science and the Journal of the Linnean Society. Although the reports are written in dry, official language, they nonetheless reveal Ward’s creativity and energy. He conducted meticulous microscopical studies on the life history of the fungus, isolating the spores and exploring how they germinated and developed through the living leaf tissue. He placed potted coffee trees around the veranda of his house so that he could observe how the disease developed on coffee plants with different exposures to the wind and rain. He hung glass slides from coffee trees to trap airborne rust spores; he deliberately infected coffee plants placed in Wardian cases. He carefully observed how rust epidemics developed in the field. He also partnered with several coffee growers to conduct experiments on chemical control. Few cultivated plants had been subjected to this kind of systematic field work, and certainly no other tropical crop had received this kind of attention. Ward’s experimental rigor matters, then and now, because he transformed scientific and popular understandings of the rust.
The central puzzle was, as it had been with Morris, to determine the full life cycle of H. vastatrix. The phases internal to the coffee leaf were, by that point, reasonably well understood, but the phases external to the leaf had still not been settled. Morris had argued that the fungus’s life cycle included an external phase in which the fungus covered the surface of the coffee plants in a microscopic mycelial web for several months. Ward quickly cast doubt on Morris’s model. He collected samples of these mycelial threads in the field and studied them under a microscope. He concluded that these filaments were produced by four species of fungi, none of which bore any relation to H. vastatrix.57 Furthermore, none of these external mycelia connected with the internal mycelia that were definitely H. vastatrix. Based on this, Ward discarded Morris’s model and the control methods on which it was based.
To clarify the rust’s life cycle, Ward conducted experiments on living plants under tightly controlled conditions. He collected rust spores from a lesion on an infected leaf. He then placed them in droplets of water on the leaf of a healthy coffee plant housed in a Wardian case. This glass case reduced the risk that the plant could be contaminated by other fungi. He found that the spores germinated in as little as twelve hours after contact with water, and the mycelium started forming inside the leaf two or three days after that. Within two weeks, this mycelium would produce a lesion visible to the naked eye. Roughly a week after that point—three weeks after the initial infection—the lesion would start producing and releasing new spores. Under ideal conditions, the lesion could continue producing spores for five to six weeks.58 He calculated that a single lesion, produced by a single spore, could produce 150,000 new spores at a time. And a badly infected leaf could contain many lesions, which could cause the leaves to drop prematurely. Each individual spore thus carried the potential to cause tremendous damage.59 As Darwin had done on a much larger scale, Ward demonstrated the tremendous cumulative power of small biological events.
Ward’s field research shed new light on the rust’s ecology. He showed how all the seemingly mysterious phenomena of the disease could be explained by the fungus’s life cycle. To determine how the spores spread in the field, he placed sterile glass slides in various parts of the farm, on the ground and attached to trees. In a single water droplet collected this way, he found spores of fifty-one species of fungi (including H. vastatrix). His experiments with slides suggested that spores could travel up to 50 feet in a single journey. He carried out other experiments that showed how outbreaks of the rust were connected to wind and rain patterns. He had placed potted coffee trees around the veranda of his house and noted that “the plants placed on the side of the house more exposed to the wind suffered more than those that had been sheltered”60 Extending these observations to the coffee estates, he argued that “a sudden appearance of the disease is closely connected with the wind and this connection is of exactly the same nature as what we should expect if the wind blows spores about.”61 Similarly, the veranda experiment showed that water was also important to the development of the disease: plants “placed on the edges of the verandah, and kept wetter on the whole (from drip, driving rain, etc) appeared to become more diseased than the sheltered ones,” an observation that he later confirmed experimentally using coffee plants in Wardian cases.62
Using the pathogenic model of disease, Ward explained the patterns of rust outbreaks in Ceylon. He described a field that had been apparently free of the rust in April but was badly infected by June. Using temperature and rainfall records, Ward showed how rains in mid-May would have caused spores across the farm to start germinating. As expected, the first lesions in the field were observed two weeks after the rain. Ward’s report discussed a number of real-life examples, showing how each outbreak could be explained mainly in terms of how wind and water shaped the fungus’s dispersal and development. He argued that the connection between the fungus and its conditions of existence “were no more mysterious than that between the life of any organism and its environment: sow the spores of Hemileia on a proper nidus, and give them air, water, and warmth, and they germinate and flourish as do the seeds of coffee or an similar plant in damp, warm, aerated soil.”63
Based on this, Ward’s recommendations for rust control differed from the ones Morris had made. Ward argued that chemical control would only be effective under