C. C. TANG AND C. T. TANG
1964, (1):
1-16.
Although raillietiniasis is recorded in all well-known textbooks of parasitology and helminthology, the life cycle of the causative organism and its mode of transmission are still unknown. In sections dealing with epidemiology and source of infection of this cestode, the hypothesis of Blanchard that the intermediate host of the parasite is probably the cockroach (Periplaneta orientalis), is universally cited. This hypothesis based in part upon the analogy drawn from the knowledge on the life history of members of Davainea known at that time, has neither epidemiologic nor experimental grounds. Pradatsundarasar (1960) reported 6 cases of human infection with Raillietina siriraji in Bangkok,Thailand. Two of the six new cases gave a history of putting a cockroach in their mouths. Based on this inquiry, he believed that the insect is the intermediate host of this tapeworm. Since the cockroaches were only put in the mouths without being actually eaten, the evidences are not conclusive to indicate that they are the responsible intermediaries of the parasite. The problem remains, therefore, unsolved.Serveys of recent literature indicate that new cases of raillietiniasis appeared in various parts of the world. Only within the last few years reports appeared on cases from Queensland and Brisban in Australia (Baer and Sandars, 1956), Batavia in Indonesia(Bonne and Mreyen, 1950), in Manila in Philippine Islands (Stransky and Lorenzo,1960), Bangkok in Thailand (Chandler and Pradatsundarasar, 1957; Pradatsundarasar,1960), and Kyushu in Japan (Miyazaki, 1950). These indicate that the disease is especially prevalent in southeastern Asia. With regard to China, cases have been reported from Taiwan (Hsieh, Maa and Chen, 1959), Foochow and Tsinkiang, Fukien, (Tang and Lin, 1959) and Canton, Kwangtung Province (Chen, 1960).Since 1953 we have been engaged in the study of the transmission problem of R.(R.) celebensis. At first we made observations on the environment of the houses where the infected children lived. Experiments were conducted to infect various kinds of insects and other invertebrates by feeding them with gravid segments of R. (R.) celebensis either from man or from rats. The insects used consisted of' larvae of. the rat flea (Xenopsylla cheopis), and cat flea .(Ctenocephalus felis), the cockroaches (Periplaneta americana), the meal-worm larvae (Tribolium ferugineum, Alphitobius piceus O1.) the larvae of the rice moth (Pyralis farinalis), larvae of house fly (Musca domestica) etc. The result of all these experiments were negative. In 1957 we began to use ants as experimental hosts. For two to three years the experiments were unsuccessful, as we were unable to keep ants alive under laboratory conditions. In 1960 when we learned the method of keeping colonies of ants alive by adapting Fielde's glass-box method as described by Wheeler (1913), we succeeded in the experiments. We fed the ants with the gravid segments of R. (R.) celebensis, which were readily carried into the nest by worker ants. In October of the same year one worker ant from a colony, which had been experimentally infected 22 days previously, was discovered to harbour two milky white cysticercoids of R. (R.) celebensis. From June to September in 1961 two more successful experiments were recorded. In one experiment, slightly immature cysticercoids were dissected out from ants 20 days after infection. In another experiment 38 days after infection, fully mature cysticercoids were secured. Each infected ant contained from one to six cysticercoids in its abdominal cavity. The immature cysticercoids possess a smaller scolex and thicker cyst wall. The scolex can protrude outside the cyst wall when the larva was removed from the haemocoel of the ant and placed in water. The mature cysticercoids were larger. Four of them were fed to a white mouse, which unfortunately died three days after feeding, and four developing scoleces were found in its intestine.They are morphologically indistinguishable from the scolex of R. (R.) celebensis. It was observed that the gravid segments, after being carried into the nest, were fed to the larval ants. Young adult ants newly emerged from metamorphosis were found infected with the cysticercoids.The mature cysticercoid is 0.292 mm in length and 0.215 mm in width. The cyst wall of the immature cysticercoid measures about 0.025 mm, while that of the more ripened one is 0.008 mm. The fully formed scolex in the cyst has a length of 0.27 mm and a width of 0.20 mm. The rostellum is 0.065 mm in length and 0.086 mm in width.The circlet of rostellar hooks has a diameter of 0.073 mm. It is larger than the diameter of the sucker. The hooks are hammer-shaped. They are of two different lengths arranged in alternate rows. The longer hooks are 21.96 u in length, while the shorter ones 19.52 μ. The details of the excretory system can be observed in the living cysticercoid. It consists of two dorsal and two ventral longitudinal vessels, which are connected to form a net-work on both sides. A ring vessel is formed surrounding the rostellum.Two vessels, one on each side, arise from the ring vessel and extend to the fore part of the rostellum. In the mid-region of the scolex there are more than 40 oval calcareous concretions which are absent in the immature cysticercoid.The cysticercoid of R. (R.) celebensis is unique in morphological characters. Comparisons were made with other known cysticercoids of the genus Raillietina, such as R.(R.) echinobothrida and R. (R.) tetragona, the cysticercoids of which were described by Horsfall (1938). The cysticercoids of R. (R.) echinobothrida were also secured by us in Futsing district from a species of ants, so it enables us to compare the specimens of both species directly. The scolex of R. (R.) celebensis is smaller. Its circlet of rostellar hooks is larger than the sucker of the scolex, while the reverse is true in the other two species. The cysticercoids of R. (R.) echinobothrida possess minute but prominent spines on their suckers, which are not found in R. (R.) celebensis. By these morphological characters and the difference in body size, these larval forms may readily be distinguished.The intermediate hosts of R. (R.) celebensis are species of ants belonging to the genus Cardiocondyla Emery, 1869. They are tentatively identified as Cardiocondyla nuda Mayer, which is distributed in the Indo-Malayan region, as having been reported from Sikkim, Bengal, Barrackpore and Ceylon.With the discovery of intermediate host and the elucidation of the life cycle of R. (R.) celebensis, it is now possible to examine the different factors of epidemiology and the mode of transmission of rallietiniasis. Since ants of the genus Cardiocondyla may build their nests in kitchens or other places near human habitation, it can be visualized that the infection of the human and rodent hosts takes place through eating the infected ants alone or with the contaminated food. The four recorded cases from Fukien were all children. Of the 35 cases of rallietiniasis reported from different parts of the world, 88.57% of them were children and 54.28% of them were under 5 years of age. This significant characteristics of raillietiniasis, is probably associated with habit of the children. It may also be due to the lack of natural immunity which is better developed in adults.The taxonomic problem of R. (R.) celebensis and its related species is discussed.The nomenclature of African and South American forms is reviewed. The authors agree with Baer and Sandars (1956) in considering Raillietina (R.) celebensis var. paucicapsulata Meggitt and Subramanian, 1927; R. (R.) funebris Meggitt and Subram., 1927; R.(R.) garrisoni Tubangui, 1931; R. (R.) sinensis Hsu, 1935; R. (R.) murium Joyeux and Baer, 1936; R. (R.) formosana (Akashi, 1916) all as synonyms of R. (R.) celebensis (Janicki, 1902). The authors are of the opinion that R. (R.) siriraji Chandler and Pradatsundarasar, 1957 is also a synonym of R. (R.) celebensis. This cestode from Thailand,according to its description, is characterized by the shape of its rostellar hooks. It was found in the present investigation that the shape of the rostellar hooks appears differently in different angles of observation. The hooks in the intact rostellum look much thinner than those dissected out from the scolex and laid flat. As shown by the drawing made by Chandler and Pradatsundarasar (1957), the rostellar hooks of R. (R.) siriraji are morphologically indistinguishable from those of R. (R.) celebensis.