114 



SCIENCE 



[N. S. Vol. XXVII. No. 081 



present, althougli its points of attachment are 

 not clear. 



t. d. a. cockerell 

 University of Colobado 



the sorghum midge 



For many years growere of sorghum in our 

 southern states have noted a common faihire 

 to produce a full crop of seed. This trouble 

 is generally known as " blast." Many scat- 

 tered references to this condition may be 

 found in agricultural literature and in cor- 

 respondence. Different theories as to the 

 cause of this observed sterility have been ad- 

 vanced from time to time. Chief among the 

 agencies which have been regarded as possible 

 causes are fungi, insects and various meteor- 

 ological conditions, such as excessive precipi- 

 tation, high humidity, severe drouth and hot 

 winds. Among the growers themselves, it is 

 quite generally held that this sterility re- 

 sults from the washing away of the pollen 

 by heavy rains during the blooming period. 



In the spring of 1907, experiments were 

 planned to determine the cause of sterility. 

 The work was largely done at Baton Eouge, 

 La., and San Antonio, Texas. In Louisiana, 

 the writer was assisted by Professor H. E. 

 Fulton, pathologist of the Agricultural Ex- 

 periment Station, and in Texas by Mr. F. B. 

 Headley, superintendent of the U. S. Ex- 

 perimental Farm at that place. Many of the 

 data on the life history are due to the studies 

 of Professor Fulton. 



The first examination of the plants at 

 Baton Eouge, made by the writer late in 

 July, disclosed the presence of large numbers 

 of a small fly on and around the heads. These 

 insects proved to be females actively engaged 

 in depositing eggs within the fertile spikelets. 

 A search of the literature available in the 

 library of the Louisiana Crop Pest Commis- 

 sion brought to light the publication by 

 Ooquillet^ of a new species of Cecidomyiid, 

 Diplosis sorghicola, received in sorghum heads 



' Coquillet, D. W., " A Cecidomyiid Injurious 

 to the Seeds of Sorglium," U. S. Dept. Agr., 

 Div. of Entomology, Bulletm (New Series), 18: 

 81-82, 1898. 



in 1895 and again in 1898. A comparison of 

 the insects in hand with this description 

 proved them identical with CoquiUet's species. 

 This identity was subsequently confirmed by 

 Professor F. M. "Webster. 



Observation showed that egg-laying began 

 as soon as the tips of the heads emerged from 

 the boot or upper leaf sheath, and continued 

 until the flowering period was wholly past. 

 Heads in every stage of development, from 

 the beginning of emergence to the close of 

 anthesis, were protected from the midges by 

 means of paper bags. The results of all bag- 

 ging experiments were in substantial accord. 

 Heads protected from the midge were uni- 

 formly fertile, where normal growth con- 

 tinued. Heads exposed during the flrst half 

 of anthesis and then protected were sterile in 

 the upper portion and well seeded below. 

 Heads exposed until flowering was wholly 

 completed were uniformly sterile when 

 midges were abundant during anthesis, and 

 partly fertile when midges were scarce. From 

 500 to 1,160 midges were hatched from each 

 of several infested heads. Exposed heads ex- 

 amined a few days after anthesis showed liv- 

 ing larvas lying in close contact with the 

 shrunken and undeveloped ovaries. The in- 

 jury is due to the absorption of the juices 

 from the young and tender ovary, thus causing 

 development to cease. This absorption is 

 through the body walls of the larva, the ovary 

 being not eaten or otherwise injured though 

 oftentimes discolored. This method of ob- 

 taining nourishment is shared also by the 

 larvfE of the closely related wheat midge, 

 Diplosis tritici, and by those of the well known 

 Hessian fly. Numerous additional facts con- 

 cerning the habits and life history of the 

 midge will be presented in a more appropriate 

 place. 



Sterility or failure to produce seed in vari- 

 ous sorghums is thus shown to be due to the 

 attacks of the Cecidomyiid midge, Diplosis 

 sorghicola Coquillet, for which the name, 

 sorghum midge, is here proposed. At Baton 

 Eouge the midge was accompanied by a 

 Chalcid parasite, a species of Aprostocetus. 

 According to Professor Webster, this parasite- 



