THE MELON ELY IN HAWAII. 45 



required by the slower vessels to cover the 2,000 miles between the 

 Hawaiian Islands and California, since he intercepted it at least once 

 a year during 1912, 1913, 1914, and 1915. Data regarding the dura- 

 tion of the immature stages show that host fruits taken on board 

 ships as ship's stores are capable of carrying the melon fly, either as 

 larvae or later as pupae in the fruit containers, for voyages much 

 longer than that required to cross the Pacific Ocean; or, in other 

 words, are capable of being the means of spreading the pest in vessels 

 plying between almost all countries where climatic conditions are 

 favorable for the establishment of the fly. There is no doubt that the 

 melon fly was carried to the Hawaiian Islands in host fruits taken on 

 ships at Japan or China, and later spread from Honolulu to all the 

 islands of the Hawaiian group in a like manner through the medium 

 of inter-island boats. 



While the spread, from one country to another at a considerable 

 distance, probably starts with the fly in the larval stage, the spread 

 from town to town, or over short distances, as between islands of an 

 archipelago, may occur in the adult or pupal stage. The writers have 

 seen female flies alight on an automobile top and be carried 16 miles, 

 from the country into the city of Honolulu. On another occasion an 

 adult was seen flying about an inter-island boat en route from Hono- 

 lulu to Hilo, on the island of Hawaii. This fly was not observed 

 after the boat weighed anchor at the port of Lahaina on the island 

 of Maui, or 72 miles from Honolulu. These two instances will ex- 

 plain the spread of the pest, in the adult stage, about the islands of 

 Hawaii, even if it could not be transported in the larval stage. 



When larvae form their puparia on bare surfaces, and particularly 

 on a cloth surface, the puparia adhere sufficient well to make it 

 possible for them to be transported considerable distances under 

 favorable circumstances. While the writers know of no instance 

 where the melon fly has been thus spread, distribution in this fashion 

 is quite feasible and to be expected. 



CONTROL MEASURES. 



NATURAL CONTROL. 



There are no agencies at present working in the Hawaiian Islands 

 that bring about, even periodically, a very large natural reduction 

 in the abundance of melon flies. The mortality among eggs and 

 larvae is not sufficient to attract attention, except under conditions 

 considered below, and then is not effective enough to be of practical 

 value as a control measure. Were bacteria and fungous diseases at 

 work, the results of such activities would have been evident to the 

 writers during the past several years. The death rate among adult 

 melon flies must be high, or the natural increase of the pest would 

 produce far greater swarms of them than are now present in fields. 

 While Severin (45) has recorded rearing 3 specimens of the hornfly 



