Blowing Tabes.

485

Chap. 2.]

the idea occurred that if the junction of B were made to form an acuteangle with the longer part of Ä, then the whole of the aerial current mightpossibly pass out at A, since to enter B it would have very nearly toreverse its direction. The device figured at No. 204 was made to test this.(The part of A in front of the joint was inches long, which from severalexperiments we thought produced the best effect, when A was half aninch in the bore—i. e. the length of this part of the blowing tube was threetimes its diameter.) Upon trial part of the current passed into B and es-caped through the liquid, as in the preceding experiment; and even whenB was turned up in a vertical direction before entering the water, the sameeffect took place.

Various modes of uniting the pipes with the view of preventing theblast from entering the vertical one were now tried, and to ascertain theeffects produced a glass tube, three feet long and three-eighths of an inchbore, was attached to the vertical or exhausting tube of each. In No. 205a portion of B protruded into A, so as to form a partition or partial coverto the orifice. Upon blowing through A (in the direction of the arrow)the water sprung up B to the height of 12 inches, and in subsequent trialsvaried from 10 to 20 inches, according to the strength of the blast. Byconnecliüg the glass tube to the blowing end of A and then blowingthrough B, the liquid rose from 8 to 10 inches; the difference no doubtbeing caused by the current of air having had greater facilities in onepassage than in the other.

We next United two tubes at right angles, but instead of making thejoint flush within as No. 203, the upper end of B was cut obliquely, asif to form a mitred or elbow jomt. This end was inserted into the underside of A, as represented at No. 206, the open part of B facing A. Theobject of this device was to ascertain whether the convex part of the ver-tical tube within A would be sufficient to divertthe blast from entering B,while it swept over the upper edge and passed round each side. Previousto connecting the lower end of B with the glass tube we inserted it inwater, and upon blowing smartly through A, the liquid rose (10 inches)and was expelled with the air, forming a dense shower. The glass tubewas then attached, (by a slip of India rubber) and upon blowing again thewater rose, on different trials, from twenty to thirty inches. The tube Awas half-inch bore, and B three-eighths. Various experiments were madeto determine the best length of that part of A in advance of the joint: theresult was generally in favor of the extent already mentioned.

The end of B cut obliquely, as in the preceding experiment, was nowinserted into A at an acute angle. See No. 207. The ascent of the liquidin several trials varied from 20 to 28 inches. A moderate puff raised it14 inches, but a strong effort of the lungs was required to elevate it overtwo feet. When the glass tube was connected to A, as in No. 208, and ablast directed through B, the highest ränge of the liquid was nine inches.

The tubes were next United as in No. 209; that is, the axis of thepart of B which entered A coincided with that of the latter, thus leavingan annular space one-eighth of an inch wide for the passage of the blast.The effect of this did not differ so much from No. 208 as was expected.The rise varied from 20 to 30 inches; and not more than half the formeramount was produced byreversing the tubes, as in No. 210. The annularpassage for the blast in No. 209 was too small, the current was pinched inpassing, and its velocity consequently diminished. In another tube inwhich the space was enlarged, the water rose six inches higher.

W e next endeavored to ascertain the effects of varying the form of thediscnarging ends of the blowing tubes, either by adapting additional ones