THE ARGENTINE ANT IN RELATION TO CITRUS- GROVES. 45 



plants, and as in the beginning it had increased more slowly on 

 the ant-free plant, it now decreased more rapidly on that plant. 



On April 25 the number of aphids on the ant-attended plant had 

 decreased to 2,043, on the ant-free plant to 802, and, from that date 

 on the decrease continued as follows: On the ant-attended plant 

 there remained, on May 6, 182 aphids ; on May 20, 18 ; on June 3, 23 ; 

 while on the ant-free plant, on May 6, there were only 52 aphids, 

 and on May 20 all had disappeared. The ants, therefore, appeared 

 to give a slight advantage to the aphids up to this time, but by 

 June 16 all had disappeared from both plants, the parasites having 

 won in the struggle for their possession. It is seen therefore that 

 although the ant attends the orange and certain other species of 

 aphids having very efficient internal parasites, it is unable to pre- 

 vent the destruction of these aphids and cause any noteworthy in- 

 crease in their number. 



TRANSPORTATION OF APHIDS AND COCCIDS. 



Although at times nearly 50 per cent of the ants foraging in citrus 

 and some other trees capture insects and carry them down the tree, 

 taking it throughout the season the average is less than 1 per cent. 

 About 8 per cent of the total foraging workers counted in all ex- 

 aminations were engaged in carrying all kinds of insects, but this 

 is, of course, above the average, as counts were made only upon trees 

 where the ants were engaged conspicuously in the transportation of 

 insects. Only a fraction of 1 per cent, viz, 0.5 per cent, of the in- 

 sects carried by the ants were scales and aphids, and only under ex- 

 ceptional circumstances is the number carried worth considering. 



Extended observations on the activity of the ants in transporting 

 scales and aphids have led to the following conclusions: 



(1) The ants feed to a slight extent upon the surplus insects when 

 its host scales or aphids are very numerous, upon those that have 

 died from parasitism or some other cause, and upon male scales as 

 fully as their ability to capture them allows. 



(2) The ants utilize dead shells of the black and some of the 

 armored scales for the construction of shelters and feed upon the 

 softer by-products and detritus of these scales. 



(3) Direct dissemination of orange scales and aphids by the ant 

 is only incidental and is negligible. Indirectly the ants aid in the 

 dissemination of some of these insects by greatly increasing them on 

 particular trees, and from these points of heavy infestation they 

 spread by the usual means. 



Some of the facts which lead to the foregoing conclusions are as 

 follows: A large majority of the scales and aphids carried by the 



