Accounts and photographs of the events indicate that although the 8 May eruption was extremely lethal, it was the 20 May eruption that carried greater destructive power. Heilprin [1903, pp. 27-29] gives the following account:
The aspect of the ruined city as I found it at the time of my first visit differed considerably from that immediately following the 8th of May and had manifestly been largely shaped by the eruption of May 20. After its first destruction, although the extinguishment of life was complete, rows of houses were left standing almost intact, notably in the central quarters of the city. Photographs taken several days after the catastrophe plainly show this feature ... Many roofs were still in position, the massive building of the mayoralty carried its overhanging cornice, and the Hopital Militaire its walled (how historic) clock. Many signs remained on the buildings, and there were other evidences of an only recently passed activity. At the later day, all this had changed.The second blast, that in intensity was nearly, if not fully, the equal of the first, laid to ground what still remained high, and gave to the city that distinctive oriented aspect which it now presents. The greater number of the massive walls run parallel with the sea, or in line to the volcano; and there are few that have been preserved in their full height that take a direction at right angles to this. It would thus seem that the destroying force of the eruption of May 20 expended its main energy along a north and south line, shattering everything that was more directly opposed to its course. This was not so markedly the case on May 8, when much of the force was directed radially.
This description raises points concerning the relative kinetic energy of the two surges, although there is some difficulty in verifying the description. The photographs that he mentions are essential since Heilprin did not arrive in St. Pierre until 25 May 1902, and his book  does not include these photographs. Nonetheless, there is some evidence to add credibility to his observation that buildings which survived the May 8 eruption were later destroyed by the 20 May eruption. The image below is a from a photograph taken on 11 May, 1902, showing victims burned by the heat of the pyroclastic surge at St. Pierre's Rue Lonchamps.
|Rue Lonchamps three days after the 8 May eruption. Note the burn victims and the very thin deposit of ash. [LaCroix 1904, p. 294].|
Rue Lonchamps is in the central area of the city, as shown on the maps below.
|A Map of St. Pierre. [LaCroix 1904, p. 229].|
|A detail of the map above, showing the central and southern portions of the city. Rue Lonchamps is highlighted. Note the Morne Abel to the north of Rue Lonchamps, which is apparently in the background of the 11 May photo. [LaCroix 1904, p. 229].|
The 11 May 1902 photo of Rue Lonchamps shows that most of the building walls are intact. Although the buildings appear to have lost their roofs, the gable walls running perpendicular to the street are intact (this is particularly clear in the buildings in the background in the upper part of the photo). In contrast, the image below shows a close-up of the vicinity of Rue de Lonchamps, extracted from the overall photograph taken in 1904. The detail image is very grainy because of the half-tone printing process, but it appears that there are few standing building facades, at least not to the degree shown in the 11 May photo. In relating the May 11 photo and the June 1904 photo, note the presence of Morne Abel, the ridge just to the north of Rue Lonchamps, which is apparently in the background of the 11 May photo. This evidence is far from conclusive, but does lend credence to Heilprin's characterization of the two eruptions.
|The overall view of St. Pierre photographed in June 1904, with a frame indicating the detail image shown below. Referring to the map shown above it is clear that this photo was take from Morne d'Orange at the south end of the city looking north toward Morne Abel, which is at the center of the detail image. [LaCroix 1904,p. 233].|
|A detail of the vicinity of Rue Lonchamps. The photographic contrast has been heightened to reveal image features. Although the detail image is not very clear, it appears that the degree of damage is much greater than that shown in the May 11 photo above. [LaCroix 1904, p. 233].|
Modern descriptions of the 1902 eruption of Pelée emphasize the destructive power of the pyroclastic surge [e.g. MacDonald 1972, p. 145; Sigurdsson 1982, p. 49], however Heilprin's account reveals another aspect of the events, showing that it is possible for a surge to be extremely widespread and lethal without being completely destructive. The temperature of the 8 May surge was extremely deadly, but its density and velocity did not result in sufficient kinetic energy to flatten everything in its path. This property is revealed in an account of one of the few known survivors of the 8 May eruption [Heilprin 1903, p. 119-120].
On the 8th of May, about eight 'o clock of the morning, I was seated on the door-step of my house. ... All of a sudden I felt a terrible wind blowing, the earth began to tremble, and the sky suddenly became dark. I turned to go into the house, made with great difficulty the three or four steps that separated me from my room, and felt my arms and legs burning, also my body. I dropped upon a table. At this moment four others sought refuge in my room crying and writhing with pain. although their garments showed no sign of having been touched by flame. At the end of ten minutes, one of these, the young Delavaud girl, aged about ten years, fell dead; the others left. I then got up and went into another room, where I found the father Delavaud, still clothed and lying on the bed, dead. He was purple and inflated, but the clothing was intact. ... Crazed and almost overcome. I threw myself upon a bed, inert and awaiting death. My senses returned to me in perhaps an hour. when I beheld the roof burning. With sufficient strength left, my legs bleeding and covered with burns. I ran to Fonds-Saint-Denis, six kilometers from St. Pierre. With the exception of the persons of whom I have spoken, I heard no human cries ; I experienced no degree of suffocation, and it was only air that was lacking to me. But it was burning. There were neither ashes nor mud. The entire city was aflame.
In this grisly account, temperature is clearly the primary effect of the surge, sufficient to fatally burn people inside their clothing, yet the kinetic energy did not collapse the building or even remove its roof, which must have been mostly in place since it was on fire.
In comparing the kinetic energy of the two events, it is important to bear in the mind that although the 8 May eruption apparently left many buildings standing, it may also have inflicted structural damage that made some buildings more vulnerable to the second eruption on 20 May. Nonetheless, it seems likely that much of the destruction of St. Pierre was the result of the 20 May eruption rather than the 8 May eruption. Considering these events with respect to the masonry walls of Pompeii and the 79 eruption of Vesuvius, the key point is that the kinetic energy of a major pyroclastic surge is not necessarily sufficient to flatten every structure in its path.
|Rue Victor Hugo, 14 May 1902, apparently looking north with Morne Abel in the background. Note the much greater degree of damage to walls perpendicular to the street. [Heilprin 1903, p. 38].|
It is important to note that this photograph was taken on 14 May 1902, and so reflects damage solely from the 8 May surge, underscoring the need to qualify Heilprin's account of limited damage due to the 8 May eruption; clearly there were areas that sustained significant structural damage on 8 May. That qualification aside, the photograph illustrates the directional nature of the loading, particularly for the building in the left foreground. The street facade wall stands a full two stories, while the perpendicular side wall stands only about one-half story height. In this condition, the facade wall has very little support in the out of plane direction and would be very vulnerable to out-of-plane loading. The fact that it is standing indicates that the alignment of the loading was very close to the alignment of the wall.
The effects of wall alignment can also be seen in the overall view from the 1904 photograph, which shows an abrupt change between an area of nearly complete destruction and an area of partial destruction to the south. The image below includes annotations which mark the alignment of Rue de Victor Hugo, the main street of St. Pierre, and the approximate boundary between the areas of complete and partial destruction. It is clear that the boundary occurs at a slight kink in the alignment of the street, marking a shift in the city grid, implying that the degree of building damage was sensitive to the building alignment.
|Rue Victor Hugo, 14 May 1902, apparently looking north with Morne Abel in the background. Note the much greater degree of damage to walls perpendicular to the street. [LaCroix 1904, p. 233].|
The shift in the grid is even more apparent in the image below, taken nearly five years after the eruption. The slightly different photo angle and the vegetation overgrowth clearly delineate the street orientation and the remaining walls.
|Overall view of St. Pierre, April 1907. The overgrowth highlights the shift in orientation of the main street. [LaCroix 1908, p. 28].|
Of course, it is important to bear in mind that alignment is not the only factor in the degree of damage. The partially damaged buildings are to the south, farther from the volcano crater, in an area where the kinetic energy of the surge would have been diminishing as it travelled along level ground. Nonetheless, the fact that the transition from complete to partial damage is so sharp at the point where the alignment changes suggests that alignment was a significant factor.
For buildings north of the transition point, it is possible that walls in both directions received an out-of-plane component of loading sufficient to collapse the wall, or, more likely, that walls aligned primarily perpendicular to the surge were destroyed by the strong out-of-plane component, allowing the relatively small out-of-plane component on the nearly parallel walls to topple those walls in turn. To the south, in the region of partial destruction, the walls perpendicular to the load were destroyed by the out-of-plane component, as shown in the scene of Rue de Victor Hugo above, but the walls parallel to the load were so closely aligned that they received negligible out-of-plane loading and were able to stand even though their out-of-plane support had been removed.
Although the damage at St. Pierre indicates that the pyroclastic surge was strongly oriented in the direction parallel to the standing walls among the ruins, it is important to recognize that pyroclastic flow and surge are turbulent phenomena where the flow direction may have strong local variations. Based on ash deposits near north of St. Pierre, Fisher  concluded that there were local areas where the flow refracted around corners and actually moved opposite the direction of the main flow. The figure below diagrams this phenomenon, which was observed in the Fond Cannonville area (see the isopach diagram, north of the main ash-and-block flow), where the flow turned the corner and proceeded up a small valley.
|A diagram of refraction of pyroclastic surge [Fisher 1982, p. 369].|
In considering the issue of flow direction with respect to phenomena at Pompeii, the oriented nature of building damage in the south of St. Pierre implies that it is important to consider global patterns of wall damage with respect to wall alignment and potential directions of the pyroclastic surge. Any conclusions, however, must recognize the potential for local variations in the flow direction.
|Next: Mt Pelée: Construction Types at St. Pierre|