A Thunder of Cannon
Archeology of the Mexican-American War Battlefield of Palo Alto
NPS Logo


As noted in Chapter 1, the primary goal of the 1992-1993 investigation of Palo Alto battlefield was identification of the battle line positions taken by the two opposing armies. Achieving this goal required four, mutually supportive sub-goals. These sub-goals and their summary conclusions are as follows:

1) Acquisition and synthesis of primary documents to provide a basis for comparison with the archeological data.

A researcher of an historical event obtains documents sufficient for plausible reconstruction of that event. A possible corollary to any effort of this scope is the concern not enough time was invested to find all of the relevant documents on which such a reconstruction must depend. Undoubtedly, great numbers of undiscovered primary sources pertaining to the battle of Palo Alto exist. Yet, given the constraints of this project, a respectable number of historically significant primary sources were obtained, analyzed and synthesized.

An attempt was made to classify primary sources as to their degrees of reliability, the resulting data synthesized and compared with archeological data. This research approach was found appropriate since: 1) archeological field time could be spent in areas where a specific battle event likely occurred; and 2) it lessened the possibility of comparing archeological data with more or less imprecise historical data. Both written and schematic battle accounts from reliable sources provided source materials for interpreting specific artifact patterns. For example, concentrations of U.S. canister shot found along the eastern margins of Area F is believed the result of the documented cannonading by Duncan's battery on the Mexican right flank.

Of course, in a perfect world, one realizes a consistent correlation between presumed "good" historical data and the archeological pattern. In fact, this desired outcome was not realized for several critical areas of the battlefield. Very little archeological evidence presently exists to support historical documentation, reliable or otherwise, regarding: 1) initial battle line locations of either army; 2) flanking and counter-flanking activities on the west side of the battlefield; and 3) location of the U.S. army during the final phase of the battle. However, negative information, that is, the absence of artifacts where in theory they should exist in relative abundance, does not necessarily throw the historical accounts into question since relic collecting and ground disturbances have resulted in a flawed archeological record for the battlefield. In addition, randomly placed survey units may have simply missed such evidence. Obviously, further fieldwork is required within these and other areas of the battlefield, guided as always by data from documentary research.

2) A study of present-day topographic features at Palo Alto will determine if these same features existed in 1846 and, if so, assess their relevance to battle conduct.

Comparisons of eyewitnesses' schematic topographic descriptions with present-day topographic features often resulted in positive correlations, actual physical links with the historical past. Thus, certain natural and cultural features found today on the battlefield were used as points of reference in choosing promising areas for archeological study. For example, local lore identifies a particular present-day low rise as "Arista Hill", one of Berlandier's motitas (Area E) that was the anchor of the Mexican right flank. If this identification is correct, then physical evidence of at least a portion of the Mexican battle line should be somewhere west of Area E. Survey of this area did, in fact, locate strong linear evidence of the Mexican battle line, presumably of the final phase of the battle. One can thus accept with confidence the local belief that "Arista Hill" is the same motita described by Berlandier. Area F is most likely the marshy area noted by a few battle participants as an obstacle between the opposing battle lines. This feature is apparent on the aerial photograph (Figure 3), and is partially delineated by an artifact distribution pattern around it (Figure 41). The biological survey of the battlefield indicates Area F extends along a roughly northeast-southwest axis, which roughly corresponds with the presumed axis of the final Mexican battle line.

The lomas tendidas of Area G seemingly lack the more obvious physical characteristics present at the time of battle. Presumably these lomas comprise a resaca levee extending in a southeasterly direction from "Arista Hill". Post-1846 erosional actions, such as from hurricane flooding, certainly could have obliterated this feature along with the vegetation that may have stabilized it. At present, heavy rains periodically delineate the existing meager remnants of this levee.

Local tradition also identified the approximate location of Palo Alto pond (Area A). Both vegetation removal and drainage activities during at least the last 50 years apparently destroyed any remaining in situ battlefield-related evidence, such as the earthen fortifications that protected the U.S. wagon train following the battle. However, a few artifacts were found in Area A, such as an ox shoe, cartridge box or knapsack rivets, and a U.S. officer's sword buckle. These artifacts lend some scanty support to the belief this area of the battlefield was associated with the initial U.S. battle line.

Area C encompassed the resaca that figured prominently during the initial battle activities. Identification is based not so much on archeological evidence as on the evidence the present-day resaca is within its documented location, and it essentially has the same meandering configuration as the resaca described in 1846. Thick vegetation cover west of the resaca prevents conclusive archeological identification of battle activities here.

Analysis of the battlefield aerial photograph hints at vestiges of the old Matamoros wagon road (Area B). Identification of disturbance vegetation dominance within Area B lends support to this belief. However, Area B also is devoid of archeological evidence. Does this mean little or no battle-related activities took place along this remaining road segment? Or is the ca. 1846 road actually somewhere else? A survey method appropriate for the thick vegetation cover is required for this area of the battlefield.

3) The post-battle history of Palo Alto indicates to what degrees the battlefield topography and artifact patternings have been affected.

There is sufficient documentation to indicate unevenly distributed adverse effects on the archeological record. Battle-related objects were removed as soon as the battle was over, both as souvenirs and as salvage for reuse. By itself, this initial removal would not have destroyed the overall artifact patterning. Unfortunately, relic collecting as a sport has damaged several nonvegetated areas easily accessible to collectors, and the greatest damage to site integrity has been caused by those using metal detectors. Site damage is largely concentrated within areas of sparse vegetation cover and where artifacts are periodically exposed due to plowing. Battle interpretation also may be compromised by those who have selectively collected only certain categories of artifacts such as Mexican accouterments, leaving behind a skewed artifact pattern consisting of mostly "rejected" artifacts, such as iron shrapnel fragments. Fortunately, a covering of tall cordgrass protects key portions of the battlefield. These areas are something of an archeological preserve, representing the last remnants of pristine artifact patterning, and must be protected from any future illicit collecting.

Not all effects on the battlefield have been deleterious. Specifically, the heavy growth of mesquite-cordgrass chaparral west of Area C, the resaca, presently protects this portion of Palo Alto battlefield. This forest has existed here for decades; therefore, it is likely relatively pristine archeological evidence of Mexican flanking maneuvers and U.S. defensive counter measures still exist below this protective vegetative covering. The same type of vegetative protection situation also may protect locations of the Mexican hospital and baggage train. Again, a method for subsurface archeological data retrieval specific to dense vegetation cover is required.

4) Analyses of individual artifacts and artifact distribution patterns may provide corroboration or refutation of specific battlefield events.

Besides simple identification of the extent and nationality of the battle lines, artifact spatial analysis provided strong physical evidence of specific battle events and was largely successful in corroborating various eyewitness accounts. Spatial analysis identified the following battlefield actions:

1) A general clustering of Mexican cannon balls occurs in one area of the battlefield where documents place Duncan's battery. This confirms battle accounts describing Mexican attempts to destroy U.S. artillery via artillery barrages;

2) Concentrations of Mexican shot probably define the area where it was documented U.S. dragoons were repulsed by this type of ordnance;

3) U.S. light artillery batteries concentrated their firepower on the closest Mexican units. This is evidenced by significant concentrations of U.S. 6-pounder canister shot between battle lines. In contrast, 18-pounder shell and spherical case shrapnel predominated within areas of the Mexican battle line beyond normal range of U.S. light artillery when firing canister;

4) U.S. spherical case shrapnel sharply delineates a segment of the final Mexican battle line position. This segment, more than 400 ft long, extends along an almost north-south axis. A number of battle maps likewise illustrate this same axis for the final Mexican battle line. This patterning is especially significant in light of the fact that previous interpretations of Palo Alto extend the battle lines along an initial east-west axis, with the second phase pivoting no more than 35° towards the northwest;

5) The greatest concentration of recorded artifacts center around the marshy area of Area F, and apparently reflects the actual focus of battle activity. A previous interpretation places the battle focus about .3 miles to the east of Area F, with a corresponding extension of battle lines much further to the east than what we believe to be the case. Also, the relative concentration of Mexican-related artifacts located just south of Area F is especially significant. This concentration may reflect tactical instability of the final Mexican battle line. Tacticians have long noted the tendency for soldiers under fire to almost instinctively draw closer together. This tendency is termed "crowding". Crowding grows stronger as the distance from the enemy narrows, and crowding along a battle line increases exposure to enemy fire. Good leadership on the battle line will keep men properly spaced. However, if leadership deteriorates or is poor at the onset, crowding escalates as enemy contact increases, with men pressed together for protection as extreme fear develops (Du Picq 1946; Marshall 1978; Keegan 1978; In Fox 1993:33). At Palo Alto, we note the presence of regimental insignia of the Mexican First, Fourth, Sixth and Tenth regiments occurring within a ca. three acre area. The mingling of individuals from four regiments within such a small area suggests crowding occurred along the Mexican battle line, a not surprising event given the fact that Mexican soldiers stood in static position for hours while under a galling artillery attack;

6) The broad prairie extending northwards from "Arista Hill" does contain a relatively widespread deposition of both U.S. and Mexican ordnance. Reportedly, a number of Mexican cupreous items, such as regimental badges and musket parts, were taken from this area by relic collectors; however no such items were recovered from this area as a result of the 1992, 1993 surveys. According to Plitt (1992) (see also Plitt Survey Area I, Figure 41, pocket insert), there once existed a heavy concentration of Mexican-related artifacts located approximately .3 miles east of where we believe the major battle activity took place. The author theorizes that artifacts found by relic collectors within this area of the battlefield were deposited when the Mexican army was retreating (as reported) in a southeasterly direction. Artifacts also may have been deposited here due to post-battle collecting of the dead and wounded by both sides, and perhaps due in part to post-battle Mexican encampment. Unfortunately, since relic collectors have compromised artifact patterning here, it may not be possible to better define this area; and

7) Recovered Mexican accouterments include a number of gun parts from the India Pattern musket and the Paget carbine, firearms recorded as typical general issue in the Mexican army. Also found were examples of military uniform items manufactured in Mexico as well as imported for use by the Mexican army. In addition to their value as identifiers of the Mexican battle line, such items are important towards broader research goals. For example, detailed analyses of manufacturing attributes of these objects provides some insight into the capabilities and limitations of Mexican light industries during this turbulent period of their history.

Discovery of the initial Mexican battle line should have been evidenced by a strong artifact patterning along a more or less east-west axis. In fact, no such patterning emerged as a result of this survey. Possible definition of this battle line may only result once surveys are allowed on property south of the 1992-1993 project areas. Little archeological evidence exists for the U.S. battle lines. The author attributes this negative finding largely to the fact the U.S. army experienced relatively little loss of men and equipment. Also, extensive land modification activities may have destroyed whatever artifact patterns existed there at one time. Nevertheless, one can make a good argument as to where the U.S. regiments were positioned largely based on documentary data, in association with provenience data from the few recovered U.S. artifacts.

Finally, evidence of flanking activity on the west end of the battlefield was not found. Probably such evidence does exist but is presently inaccessible due to heavy mesquite cover. Significantly, a local informant stated several Mexican Light Infantry belt buckles were recovered by a relic collector who favored working along the southern margins of this mesquite forest. Our survey of this particular area failed to produce any artifacts, possibly because the reported collector was thorough. Although anecdotal in nature, such information underscores the ongoing problem of battlefield depredation.


This most recent archeological study of Palo Alto resulted in the survey of approximately 8 percent of the entire 3,400 acre area of the park, with emphasis placed on sampling the approximate core battle area. These surveys indicate future field investigations should continue to employ a sample approach utilizing metal detectors. Use of metal detectors by skilled personnel is the most efficient survey approach, given that Palo Alto possesses a virtual absence of surface artifacts. Also, the dense vegetation cover inhibits the employment of other, conventional survey methods dependent entirely on visual surface inspections and shovel tests. The planning of future surveys within heavily wooded areas must consider the soundness of removing extensive stands of mesquite needed to conduct a statistically valid sample. One must decide whether such removal and consequent detriment to the environment is worth the large risk that nothing of consequence will result from its survey.

The battlefield undoubtedly is the single largest and most significant cultural resource of Palo Alto. However, one should not forget other cultural resources not directly related to this battle. Specifically, future research might consider focusing on identifying "Worth's Camp", the American army encampment of March 24-27, 1846 and located just east of Palo Alto pond. Worth's Camp may, or may not, be same location of the Mexican army that bivouacked at Palo Alto on May 3-5, 1846. This area could have been obliterated by land levelling activities; nevertheless, further research is warranted in this area of the park to make certain this is the case. Neither do we know where the Mexican army positioned its field hospital and baggage train, nor where it encamped the night following the battle. Also, Palo Alto was an encampment site for American volunteers and regular soldiers both during and after the war. We know the approximate location of Palo Alto House, but we do not know the extent or complexity of the nascent community that began to develop around it until the community was abandoned in the early 1850s. Palo Alto is believed to have been home to a few homesteads; these need to be located and recorded as well.

Most important, a thorough historical documents research program addressing such research goals must be conducted prior to initiating any additional archeology within the park. Without the guidance of such documentation, archeological fieldwork becomes a wasteful lesson in futility.

During the Mexican-American War only two major battles took place within the present boundaries of the United States: Palo Alto and Resaca de la Palma. Other engagements occurred in Texas, New Mexico and California but were little more than skirmishes although significant if only on a local level. As a result of urban growth, Resaca de la Palma battlefield has virtually disappeared as a cultural resource, thus bestowing singular importance on the remaining battlefield located just up the road. In this light, "future Mexican-American War studies" would seem somewhat grandiose if necessarily limited to Palo Alto alone. Yet, during this war, over 20 major engagements occurred, most of which eclipsed Palo Alto in numbers of participants, areal extent of fields of operations, and complexities of applied tactics and logistics. These major battles all took place south of the Rio Grande. Therefore, it is indeed fortunate both Mexico and the United States are working toward the preservation of their respective Mexican-American War battle sites.

While this study has many purposes, one purpose it definitely does not serve is the final and incontrovertible explanation of what happened on the prairie of Palo Alto during the afternoon hours of May 8, 1846. There is ample room for additional research and analyses which, in turn, could lead to refinement, even radical reconstruction, of the hypothetical reconstructions presented here. This study provides a framework for further research; it is a starting point, not a conclusion.

<<< Previous <<< Contents>>> Next >>>

Last Updated: 25-Feb-2009