days were theoretically as much a part of the ancient fast system of the Church as Lent, they had at that time no place whatever in the Prayer Book, and Becon had to explain to his readers the generally received reason of the name. "By opinion of much people these days are called Imber days because that our elder fathers would on these days eat no bread but cakes made under ashes, so that by the eating of that they reduced into their mind that they were but ashes and so should turn again and wist not how soon" (Reliques of Rome, sec. 58, Works, 1563, pt. iii., ff., 353). During all the medieval period the word had survived in popular speech among the Church-words that have reached us in a Saxon dress, and that the generation which called their fast "embering days" were thinking of burnt or burning wood when they thought at all, should hardly be doubted. Becon's testimony is sufficient, while the word too was plausible. There was the analogy of Ash Wednesday to keep the idea alive, not to speak of the circumstance that the very Wednesday after that day was the first Ember day of the year.

66

The earliest official recognition of the word by the English Church was in 1604, in the 31st Canon, wherein occurs the passage, Jejunia quatuor temporum, commonly called 'Ember-weeks."" The first entry of the word into the Book of Common Prayer was in the Scottish edition of 1637, where the Ember-week Collect and rubric occurred, though Ember had no place in the Tables. Ember then meant burnt or burning wood; it was the very period of Milton's "glowing embers through the room," and the word must have meant that exclusively to every ear and on every tongue. In 1657 Sparrow quoted Becon's traditionary explanation of Emberdays with approval. L'Estrange in 1659 did not touch on the English word. In 1662 the English Common Prayer for the first time admitted "Ember," both in the rubric and in the Tables, and the word then bore but one meaning, Milton's meaning. Whether, however, this was the meaning of the old Saxon word with which we started is quite another question. At length, in 1665, just after we were committed to the word, it came in the way of the learned Dr. Thomas Marshall, while labouring as a Saxonist and annotating the ancient Gospel already spoken of, p. 60 of the volume Quatuor Evan

geliorum versiones perantiquæ (1665). In his day he found (p. 528), the word written Imber and Ember, with various opinions entertained as to its etymology, some considering it akin to the word meaning ashes, others deriving it from the German antbehren, denoting abstinence, others from the Danish Temperdage, representing the Quatuor Tempora, while one "very learned man" assigned it to the Greek μépa, in the sense of a fixed and stated day. But Dr. Marshall's studies had brought him acquainted with a variety of the very word ymbrene in other Anglo-Saxon writings, and in quite a different connection. There was, e.g., a passage containing the expression thas geares embryne, "per anni circulum," a year's course or circuit, and the Saxonist at once perceived that he had discovered the source of the original word. It was to be analysed, he saw, not as im-bren, containing the element of burn and brand, which several languages exhibit, but as ymbe- or emberyne, circum-cursus, the compound denoting a running round, or circuit. He felt, however, an explanation was called for as to why these fasting seasons in particular should have been designated by a term which was equally applicable to all seasons that regularly recur, and he suggested that those in question were at first variable and arbitrary, but eventually became fixed and stated, and therefore were described as coming round, or in course. It might possibly have suited him better to recollect the fact that the quatuor tempora were from the very first linked with the four seasons of the year, spring, summer, autumn, winter, and that the expression "revolving seasons has ever been a favourite one. But whatever was the rationale of ymbrene to express what it was intended for, it was too abstract and loose for the purpose, and a familiar symbol, carrying a specious and obvious meaning, succeeded in running away with and retaining possession of the word.

In 1672 Dr. Comber, the next great expositor after the appearance of Marshall's note, made no remark upon the word, and the writer to popularise that scholar's conclusion was Robert Nelsor, who called attention to and declared in favour of it in his "Fasts and Festivals," 1704. Nicholls however, in 1710, did not notice it, and explained ember by pépa, without further remark. In 1755 appeared Dr. Johnson's

Dictionary, which, without any decision of its own, directed the reader to Nelson. Mant likewise, in 1820, while recording both the old opinion and the new one, reserved his own. Perhaps the reason why the learned were so slow to close with Marshall may have been the apparently far-fetched etymology proposed for so common and fixed a word as ember. Later writers, however, have had no misgiving, and generally agree to dismiss every idea of ashes without even so much as a hearing. Yet it may be questioned whether any amount of philological handling will help ember or make it plainer than it is. What the learned, Marshall included, have explained is not really ember; it is ymbrene or imbren. What they may have proved in regard to ember is that the word has no right to the place it occupies and is not the true descendant of ymbrene. Ember was put into the Prayer Book at a time when the primitive word was buried, and it is difficult to see how any amount of etymology can now convert it to Imbren.

EVENSONG (Vespertinæ Preces and Vesperæ in the Latin Prayer Book of 1560), the same as Vespers, and so employed in the Primer of 1410 (Mask. M. R. iii. 62). Evensong was made Evening Prayer in 1552, but was not then augmented with the penitential opening as Morning Prayer was. That opening was first prefixed, and the concluding prayers first added, in the Scotch Prayer Book of 1637, and in the English one in 1662. See more under MATTINS.

GOLDEN NUMBERS, the numbers from I. to XIX. prefixed to certain days in the English Calendar, from March 21st to April 18th, both inclusive, in order to assist in showing when Easter falls in any given year. They are based upon the fact that the moon's similar changes (full-moon for instance) occur on a different day of the Calendar through nineteen years; after which period the days for the same changes recur in precisely the same order for another nineteen years. instance, March 27th was a full-moon day in 1831, 1850, 1869, and will be again in 1888. This period of nineteen years is called the Lunar Cycle, and was employed for Calendar purposes long before the Christian era. Any year whatever might have been fixed upon to begin the Cycle with, and the

For

one which did start it was that which has made the following years, for instance, in these times the commencing ones: 1824, 1843, 1862, 1881, 1900. Thus the nineteen years, 18241842, 1843-1861, etc., are Lunar Cycles. Obviously every year is somewhere in a Lunar Cycle. Thus, since 1824 is the first year of a cycle, 1828 is the fifth; and since 1881 is another first, 1887 is the 7th. Whatever place in a cycle a year occupies that is its Golden Number; and so the Golden Number of 1828 is V., of 1887 VII. A simple rule will find the Golden Number of any year without the necessity of our remembering the commencing years of the cycle. Add 1 to the year, divide by 19, the remainder (or 19 if there be no remainder) will be the Golden Number. For 1760 it is XIII., for 1766 it is XIX. The reason for placing these numbers against certain days of the calendar, which is the next thing to be considered, is based on the definition of Easter-day, viz. that it is the Sunday after that full-moon which falls upon or next after the vernal equinox (March 21st). Now all the full-moon days of the Lunar Cycle are known by astronomy and tabulated, and therefore those which fall upon or next after March 21st, the latest of these being April 18th. These two days are the Paschal full-moon limits, and from them it follows that the earliest Easter Day is March 22nd, and the latest April 25th. Now let us revert to any one of the years which commence a Lunar Cycle, say 1824. Astronomy shows that in that year there is no full-moon on March 21st, and that the first one after occurs on April 13th. We thus obtain the formula I. April 13th, which expresses that the Paschal full-moon for the first year of the cycle is always April 13th. Again, in 1825 (or any other second year of a cycle) the earliest fullmoon after March 21st is April 2nd, giving us the formula II. April 2nd, expressing that in the second year of a Lunar Cycle the Paschal full-moon falls on April 2nd. For the third year of the cycle we get III. March 22nd. Proceeding to the end of the nineteen years we obtain the following series, in which we may notice that it is in the fourteenth year of the cycle that March 21st is a full-moon day, and in the sixth that April 18th is :-V. March 30th, VI. April 18th, VII. April 7th, VIII. March 27th, IX. April 15th, X. April 4th, XI. March 24th, XII. April 12th, XIII. April 1st, XIV. March 21st, XV.

April 9th, XVI. March 29th, XVII. April 17th, XVIII. April 6th, XIX. March 26th. These are the Paschal full-moon days arranged in the order of the Golden Numbers, exhibiting the day for each successive year of the cycle. Let us test them by the Almanac full-moons in a single year, say 1887. For 1887 the Golden Number is VII., and by the above series the Paschal full-moon should occur on April 7th; yet by the Almanac a full-moon does not arrive until 39 minutes past 5 on the morning of April 8th, a discrepancy which necessitates a few more remarks. The motions of the heavenly bodies are not absolutely equable at all points of their orbits, being sometimes faster and at other times slower than their average or mean rate. We are familiar with this fact in the case of the clock, which shows the sun's time in the main, but not in detail or according to the sun-dial. The clock is regular and undeviating, while the sun is nearly always a few minutes or seconds after it or before it, as the Almanac daily shows. The clock exhibits that time which the sun would make if it moved evenly at its mean rate, or that time which an imaginary body called by astronomers a mean sun would make. Now for Calendar purposes a mean sun and a mean moon are used, and generally speaking the full moons which these imaginary bodies produce occur on the same day as the real full moons, though at different hours, and then the Almanac and the Calendar (which notes the day only) agree; but occasionally the actual full slips a little beyond the calendar day, or else just fails to reach it, like as the sun a little outstrips or lags behind the clock, and then a discrepancy ensues, as in 1887. The Paschal full-moon is, of course, always that of the Calendar, and it is sometimes called the ecclesiastical full-moon as distinct from the astronomical one.

66

Having obtained the Paschal full-moon days in the order of the Golden Numbers, we proceed now to re-arrange them according to the days of the Calendar, and the series becomes:-XIV. March 21st, III. March 22nd, XI. March 24th, XIX. March 26th, VIII. March 27th, XVI. March 29th, V. March 30th, XIII. April 1st, II. April 2nd, X. April 4th, XVIII. April 6th, VII. April 7th, XV. April 9th, IV. April 10th, XII. April 12th, I. April 13th, IX. April 15th, XVII. April 17th, VI. April 18th. This series will be found in the