296 REPORT 4, UNITED STATES ENTOMOLOGICAL COMMISSION. 



Macliiiies of the above description but of smaller size do not need 

 the legged wheels, but are mounted entirely upon the cart or wagon, 

 in which case the cross-bar of the frame rests upon the rear part of 

 the vehicle, and the front angle of the frame is attached farther for- 

 ward ; but where extremest width is aimed at wheels so close together 

 as those of a wagon are objectionable, since their lateral rocking on 

 passing upon the row-ridges or other irregularities would cause the 

 long arms to be thrown alternately up and down to an undesirable ex- 

 tent. This can be somewhat obviated by directing the arms somewhat 

 upward, but where greatest expanse is sought wheels farther apart to 

 straddle two or more rows will remedy the fault just mentioned. As a 

 rule, smaller machines will probably be used to avoid the extra expense 

 of special wheels and legs. And where cotton is uniformly planted, or 

 the adjustability is not aimed at, the cross-bar alone, in its simplest 

 form, or a pair of the long arms spliced together to extend straight 

 across and far off from the vehicle, will, of course, be sufficient with 

 any forms of ordinary jointed tubes. Yet there are practical plans 

 for altering the pipes on the straight cross-bar to suit row-spaces of 

 diverse widths. The first is, to have no intermediate segments joining 

 the pipes but have them all radiate from a central source to the bar, on 

 which they are adjustably fixed by screw-clamps, hooks, or otherwise, 

 and from which they continue parallel onward to their juncture with 

 the distal arms. This makes a pii)e-system of the smallest number of 

 pieces, but it takes more of the tubing than does the segmental con- 

 struction. 8eco7idlyythe method of emj)loying segments which are ad- 

 justable in length can be resorted to. 



The patterns I have made for this arrangement are represented in 

 Plate XLIX, Figs. 9 and 10. We may first consider that shown in 

 Fig. 9. The metal part with its side branch is indicated by m h. The 

 other half of the segment is a piece of hose, li, clamped to the metal 

 parts by wire wraps, o, at its ends. It will be seen that the metal 

 X)art telescopes deeply into the hose and the two tubes thus overlap 

 concentrically so far that there is a wide range for the adjustment of 

 the one upon the other in a back and forth direction to set the section 

 thns formed wider or shorter as may be required. As already stated 

 above, the hose piece may have inside itself a small rod to give it sup- 

 port. The same effect is also accomplished by the design in Plate 

 XLIX, Fig. 10, where m h is a smaller half of the segment telescoped 

 into m, the larger half, and at h is a sheath-coupling wired to both at 

 00. The former of the. two plans gives a simpler and lighter set of 

 tubing. The lightest set of tubes and simplest to make is wher* only 

 hose is used for the segments and all parts except the angle- coui^lings 

 and nozzles, while the flexible parts are sustained by tension or by 

 small rods inside. The cost of materials for sets of this kind is not 

 greater than for brass tubing, though less durable. Also the cost of 

 making is not so great with the rubber j and then any planter can 



