winding just before each time of striking. And that is done by a lever being tipped over by the snail at that time, which at once stops the winding. When the striking is done the man can put the lever up again and go on. The loose winding wheels are not pumped in and out of gear as usual, being too heavy, but one end of the arbor is pushed into gear by an eccentric bush turned by the oblique handle or lever which you see near the upper corner of each striking part, and they can be turned in a moment. They are held in their place for gear by a spring catch to prevent any risk of slip ping out. Moreover the ropes themselves stop the winding when the weights came to the top, pretty much as they do in a spring clock or a watch, though not exactly.
The mode of letting off the hour striking is peculiar, with a view to the first blow of the hour being exactly at the GOth second of the GOth minute. It was found that this could not be depended on to a single beat of the pendulum, and probably it never can in any clock, by a mere snail turning in an hour, unless it was of a very inconvenient size. Therefore the common snail only lets it off par tially, and the. striking stop still rests against a lever which is not dropped but tipped up with a slight blow by another weighted lever resting on a snail on the ] 5-miiiute wheel, which moves more exactly with the escapement than the common snail lower in the train. The hammer is left on the lift, ready to fall, and it always does fall within half a second after the last beat of the pendulum at the hour. This is shown in fig. 28, where BE is the spring stop noticed above, and P the ordinary first stop on the long lifting lever PQ (which goes on far beyond the reach of this figure to the hour snail). The second or warning stop is CD, and BAS is the extra lever with its heavy end at S on the 15-minute snail. When that falls the end B tips up CD with certainty by the blow, and then the striking is free. The first, second, and third quarters begin at the proper times ; but the fourth quarter chimes begin about 20 seconds before the hour.
The clock reports its own rate to Greenwich Observatory by gal vanic action twice a day, i.e., an electric circuit is made and broken by the pressing together of certain springs at two given hours. And in this way the rate of the clock is ascertained and recorded, and the general results published by the astronomer royal in his annual report. This has been for some years so remarkably uniform, that the error has only reached 3 seconds on 3 per cent, of the days in the year, and is generally under two. He has also reported that "the rate of the clock is certain to much less than a second a week " subject to abnormal disturbances by thunder storms which sometimes amount to seven or eight seconds, and other casualties, which are easily distinguishable from the spontaneous variations. The original stipulation in 1845 was that the rate should not vary more than a second a day not a week ; and this was pronounced impossible by Mr Vulliamy and the London Company of Clock- makers, and it is true that up to that time no such rate had ever been attained by any large clock. In 1851 it was by the" above- mentioned clock, now at King s Cross Station, by means of the train remontoirc, which was then intended to be used at Westminster, but was superseded by the gravity escapement.
![Fig. 28.](http://upload.wikimedia.org/wikipedia/commons/thumb/a/aa/EB9_-_Clocks_-_Fig._28.png/250px-EB9_-_Clocks_-_Fig._28.png)
Fig. 28.
The great hour tfell, of the note E, weighs 134 tons and is 9 feet diameter and 9 inches thick. The quarter bells weigh respectively 78, 33, 26, and 21 cwt. ; with diameters 6 feet, 4J, 4, and 3 feet 9 inches, and notes B, E, F sh. and G sh. The hammers are on double levers embracing the bells, and turning on pivots pro jecting from the iron collars which carry the mushroom shaped tops of the bells. The bells, including 750 for recasting the first great bell, cost nearly 6000, and the clock 4080. The bell frame, which is of wrought iron plates, and the dials and hands, all provided by the architect, cost 11,934 a curious case of the accessories costing more than the principals.
(e. b.)
CLOISTER (Latin, claustrum ; French, cloitre ; Italian, ckiostro ; Spanish, claustro ; German, Hosier). The word " cloister," though now restricted to the four-sided enclosure, surrounded with covered ambulatories, usually attached to conventual and cathedral churches, and some times to colleges, or by a still further limitation to the ambulatories themselves, originally signified the entire monastery. In this sense it is of frequent occurrence in our earlier literature (e.g., Shakespeare, Meas. for Meas., i. 3, "This day my sister should the cloister enter"), and is still employed in poetry. The Latin claustrum, as its derivation implies, primarily denoted no more than the enclosing wall of a religious house, and then came to be used for the whole building enclosed within the wall. To this sense the German " kloster " is still limited, the covered walks, or cloister in the modern sense, being called " kloster-gang," or " kreuz-gang." In French, as with us, the word cloitre retains the double sense.
" cloister " denotes the quadrilateral area in a monastery or college of canons, round which the principal buildings are ranged, and which is usually provided with a covered way or ambulatory running all round, and affording a means of communication between the various centres of the eccle siastical life, without exposure to the weather. According to the Benedictine arrangement, which from its suitability to the requirements of monastic life was generally adopted in the West, one side of the cloister was formed by the church, the refectory occupying the side opposite to it, that the worshippers might have the least annoyance from the noise or smell of the repasts. On the eastern side the chapter house was placed, with other apartments belonging to the common life of the brethren adjacent to it, and, as a common rule, the dormitory occupied the whole of the upper story. On the opposite or western side were generally the cellarer s lodgings, with the cellars and store-houses, in which the provisions necessary for the sustenance of the confraternity were housed. In Cistercian monasteries tho western side was usually occupied by the " domus cou- versorum," or lodgings of the lay-brethren, with their day- rooms and workshops below, and dormitory above. The cloister, with its surrounding buildings, generally stood 011 the south side of the church, to secure as much sunshine as possible. A very early example of this disposition ia seen in the plan of the monastery of St Gall (Abbey, vol. i. p. 12). Local requirements, in some instances, caused the cloister to be placed to the north of the church. This is the case in the English cathedrals, formerly Bene dictine abbeys, of Canterbury, Gloucester, and Chester, as well as in that of Lincoln. Other examples of the north ward situation are at Tintern, Buildwas, and Sherborne. Although the covered ambulatories are absolutely essential to the completeness of a monastic cloister, a chief object of which was to enable the inmates to pass from one part of the monastery to another without inconvenience from rain, wind, or sun, it appears that they were sometimes wanting. The cloister at St Alban s seems to have been deficient in ambulatories till the abbacy of Robert of Gorham, 1151- 1166, when the eastern walk was erected. This, as was often the case with the earliest ambulatories, was of wood covered with a pentice roof. We learn from Osbern s account of the conflagration of the monastery of Christ Church, Canterbury, 1067, that a cloister with covered ways existed at that time, affording communication be tween the church, the dormitory, and the refectory. We learn from an early drawing of the monastery of Canter bury that this cloister was formed by an arcade of Norman
arches supported on shafts, and covered by a shud roof.