It is not specially to the quantity of material ejected from Krakatoa that it owes its reputation. Great as it was, it has been much surpassed. Professor Judd says that the great eruptions of Papapandayang, in Java, in 1772, of Skaptur Jokull, in Iceland, in 1783, and of Tamboro, in Sumbawa, in 1815, were marked by the extrusion of much larger quantities of material. The special feature of the Krakatoa eruption was its extreme violence, which flung volcanic dust to a height probably never before attained, and produced sea and air waves of an intensity unparalleled in the records of volcanic action. Judd thinks this was due to the situation of the crater, and the possible inflow through fissures of a great volume of sea water to the interior lava, the result being the sudden production of an enormous volume of steam.
The red sunsets spoken of above were so extraordinary in character that a fuller description of them seems advisable. A remarkable fact concerning them is the great rapidity with which they were disseminated to distant regions of the earth. They appeared around the entire equatorial zone in a few days after the eruption, this doubtless being due to the great rapidity with which the volcanic dust was carried by the upper air current. They were seen at Rodriguez, 3,000 miles away, on August 28, and within a week in every part of the torrid zone. From this zone they spread north and south with less rapidity. Their first appearance in Australia was on September 15th, and at the Cape of Good Hope on the 20th. On the latter day they were observed in California and the Southern United States. They were first seen in England on November 9th. Elsewhere in Europe and the United States they appeared from November 20th to 30th.
The effect lasted in some instances as long as an hour and three-quarters after sunset. In India the sun and skies assumed a greenish hue, and there was much curiosity regarding the cause of the “green sun.” Another remarkable phenomenon of this period was the great prevalence of rain during the succeeding winter. This probably was due to the same cause; that is, to the fact of the air being so filled with dust; the prevailing theory in regard to rain being that the existence of dust in the air is necessary to its fall. The vapor of the air concentrates into drops around such minute particles, the result being that where dust is absent rain cannot fall.
As regards the sunsets spoken of, there are three similar instances on record. The first of these was in the year 526, when a dry fog covered the Roman Empire with a red haze. Nothing further is known concerning it. The other instances were in the years 1783 and 1831. The former of these has been traced to the great eruption of Skaptur Jokull in that year. It lasted for several months as a pale blue haze, and occasioned so much obscurity that the sun was only visible when twelve degrees above the horizon, and then it had a blood-red appearance. Violent thunderstorms were associated with it, thus assimilating it with that of 1883. Alike in 1783 and 1831 there was a pearly, phosphorescent gleam in the atmosphere, by which small print could be read at midnight. We know nothing regarding the meteorological conditions of 1831.