As was pointed out at the beginning of this chapter, we can see perfectly two periods of Tihuanacu in the construction of Kalasasaya: the Second and the Third. We have considered before and in some detail, the works of the Second Period. We have only to discuss the constructions of this last period.
First of all, to this period belongs the balcony wall which took the place of a similar wall of the Second Period --- probably in the same location. The balcony wall also had as one of its objects, just as in this period, the screening from the eyes of the profane the mysteries of the priests when they made their solar observations to determine the seasons, festival dates and everything connected with these matters, or in other words, the calendar.

Chronologically, there belongs to this wall the building located within the great space which we call "sanctissimum", and in addition the Sun Door and the "repairs" carried out with magnificently carved blocks and set in the main east wall. These blocks are very noticeably distinguishable from the primitive ones (also of andesitic lava) of the Second Period, because of the different technique employed and the minimum of abrasion through erosion, (Fig. 13a).

In the year 1903 and the beginning of 1904, the Crequi de Monfort Mission made various excavations in Tihuanacu and in the course of one of them a cut was made from east to west more or less through the center of Kalasasaya. In this excavation there appeared the perron and later, in a trench that was opened, a quantity of magnificently carved blocks, (See Map III). Among these, now near the west wall, there appeared a great flat slab with a smooth surface.

A little farther back, or toward the east, there was found a stair with two steps in one piece and another block of longitudinal form which has a depression, a sort of indentation, which doubtless served to place in it another standing block for which the former served as a base. But the most important thing found upon opening that trench, was a block of a certain special extremely rare trachyte, (Fig. 26) obtained only many leagues from Tihuanacu, and which showed evidence of having been brought from a great distance.

This block, which is in the highest part of the temple, forms the crowning part of the external west wall of the "sanctissimum" and as such was supposed to serve, in our opinion, as the base for the Sun Door; that is to say, when the base and the Sun Door were completely finished. This block which we have designated as one of "observation" of the Third Period, for reasons which we shall study later (Fig. 26) is, naturally, only the central part of a basic wall which must have existed formerly and which today has disappeared like almost all of the component parts of the "sanctissimum." From such blocks, we repeat, they also constructed the temple to the new faith, the Catholic, in the modern village of Tihuanacu.

As can be seen clearly in the general map of Kalasasaya inserted in the present volume, and especially in the side views, the "observation block" is on its perpendicular, and is located on the highest part, and at the same time in the center of the palace. It represents the most prominent part of the construction which we have called the "sanctissimum." This building, encompassed within the enclosure of the temple, constitutes a quadrilateral which in its proportions has a LENGTH-WIDTH INDEX of 89, the index of Kalasasaya of the Second Period being 91. Of course, this small difference is no coincidence or accident as we shall see farther on. Its size is: length 71.80 meters; width 63.60 meters.

The "sanctissimum" --- a small subterranean temple ---is composed in its circumference of three terraces which in turn form three steps by means of which this construction is reduced and deepened toward the interior part. But the most interesting part of the remains of this ancient construction, among all that which is standing, is the northeast pillar, which served the same purpose as the northeast pilaster of the Second Period; that is to say, to mark the Winter Solstice, (Fig. 25).

But the real significance that this block has for our investigations is that it shows us that the priest-astronomers of the Third Period no longer observed on the corners of the blocks as happened in the calculations of their learned forebears of the Second Period, BUT AT THE CENTER OF EACH PILASTER as we shall also see farther on in the case of the balcony wall.

The andesitic pillar which we are studying (Fig. 25) demonstrates in an obvious fashion what has just been affirmed. It can be seen that it had a small window surrounded with the staircase ornament in the typical style of Tihuanacu. Thus it results that upon observing the rising of the Sun in the Winter Solstice, from the previously mentioned trachyte observation block of the Third Period, at the exact moment when the Sun appeared on the horizon or on the apparent horizon, it lighted the little window for some seconds and from the observation point there could then be noted a luminous vertical ray.

This was the case because the glance of the observer passed the little window of the block obliquely, or at angle of 24° 38'. Naturally, that pilaster was not isolated as it is today, but formed the corner of an external wall of the "sanctissimum" and had moreover, as can be seen on the drawing of this block, a superstructure which- can be studied in the reconstructed picture, (Fig. 25a).

As well as possible, with the few remains of the "sanctissimum" left by ancient and modern destroyers, we have measured its exterior dimensions, which are: seventy-two meters, ten centimeters long by sixty-four meters, twenty centimeters wide. As regards the internal terraces, the blocks which formed them, marked on the map (Vol. I, Pl. III) drawn from 1904 to 1912, have almost totally disappeared. On the new general map of Kalasasaya (Pl. III) inserted in the present volume, are marked only the blocks which exist at the present time. The terraces served simply to support the internal walls and on that account had a width of some 41 m. 90 cm. only.

Also one of the interesting blocks is that of the north external wall of the "sanctissimum", which contains an unfinished hole. It is notable that, having begun to bore the block on both faces, and by the already discovered system of two drills, WITHIN THE HOLE THERE ARE TWO LATERAL PERFORATIONS which might have served, when the working of this block was completed, to place a cross wire or diaphragm and cross wires for stellar observations. Another supposition concerning the object of these interior, lateral cracks, would be that they were to hold an auxiliary device which served as a guide for the auger in drilling the mail hole.

Other blocks, admirably carved in the center, have been found in Tihuanacu and these could have had no other object than to serve for astronomical purposes. For example, on the bridge near Km. 26 of the railroad to Guaqui, there is a completely finished block of this sort, ⁽⁸¹⁾ which has been removed from the ruins, and which still shows the cruciform marks for an alidade. This block, carved like all those which form that bridge, comes from the destruction of the monuments of Tihuanacu during the construction of this railroad.

As has been pointed out before, it is presumable that the supposed "sanctissimum" of the Third Period which is found in the enclosure of the Kalasasaya of the Second Period, may have been a subterranean temple similar to the palace or temple of the First Period, the structure of which we know to its last details, since when it was excavated it was found to be almost intact. Recently there was also found a small semisubterranean temple in Lukurmata. If this enclosure of the Third Period had been finished, there would doubtless have been placed on its deepest interior surface the most important sacred part of the worship of Tihuanacu.

This was the case with the aforementioned semisubterranean building of the First Period, on the interior plane of which, or rather, on its lower floor, there was recently discovered the enormous idol more than seven meters in height which was placed there in the First Period and later in the Second or Third period reworked to transform its rough surface. When this monolithic idol of red sandstone (Grauwake) was excavated in 1932, there were found with it other smaller sized idols done in the primitive technique of the First Period; they showed no subsequent retouchings of any sort carried out during the Third Period.

Since Kalasasaya of the Second Period was also perfected and used in the Third Period, other remains of this same period are discovered on its site, although for astronomical work and the calculation of agricultural dates and ritualistic festivals, the astronomical points of that building were no longer used; rather, there were used only the constructions for this purpose of their own period, or precisely, the astronomical points determined and accepted for the Third Period, such as the northeast and southeast corners of the "sanctissimum" for the sunrises, and the centers of the blocks of the balcony wall for the sunsets.

Without undertaking extensive and conscientious reconstructions of even new, excavations on the basis of continual and very exact cartography, it would be difficult to be more definite with regard to the site of the "sanctissimum." The only presumption that we may make is that the interior walls of the temple perhaps had, as in the temple of the First Period when it was finished, heads of human sculpture set in them. This assertion is based on the fact that there have been found on the site and outside of it, finely carved blocks in the technique and material of the Third Period, which show human faces in high relief, (Fig. 22).

From our point of view, the most important is the great rough stone of trachyte located on the west wall of the "sanctissimum" and which we have called "observation block" (Fig. 26) and which we further judge to have been previously planned as the final base for the Sun Door. This was to be located on the monumental external wall the center of which, we repeat, was without any doubt to be the definitive location of the formidable carved mass which the common people today call "Inti-Punku", or, Sun Door. Later, when we consider the astronomical science of the Third Period, we shall again consider this important remnant of the "sanctissimum".

Now it would be fitting to consider the last remains of the megalithic wall located outside Kalasasaya of the Second Period and which has been called up to this time, Collonade, Stonehenge, Balcony Wall, Peristyle, Projection, etc. Of this there remain standing today --- after the destruction suffered in the building of the temple to the new faith --- only nine great pilasters, one of which is lying on the ground. The other pilaster which is missing between the fourth and fifth block, counting from the south, is being used as the foundation for a native fence, some 242 meters to the west, (Figs. 31a and 31b).

The direction or alignment of the balcony wall, is typical of the Third Period, or, on the meridian with the insignificant variation of 42' east. That is to say, looking toward the south, the balcony wall runs 42' easterly. Because of this new orientation, it does not run completely parallel to the west wall of the Second Period, which deviates from the meridian 1° 6' 30" in the same direction, and thus there exists a difference of 24' 30" between the two periods.

The wall, of which these pilasters still standing once formed a part, was not constructed on plumb, to judge by the way it leans in. As we shall see later it varied inward from the vertical 2°.
That the pillars or columns which formed the balcony wall were in their time a true wall, is shown in the first place by the remains of a wall still to be found at the foot of the extreme south pillar. (See Vol. I, Pl. XV a).

Some few decades ago there must still have existed fragments of intermediate walls between the pilasters of the balcony wall; these were systematically carried off by the unscrupulous half-breeds of the village ⁽⁸²⁾ to carve troughs for hog food. Neither Squier nor Tschudi y Rivero found the remains of intermediary walls still on the surface as the illustrations of their works indicate, for which reason the former called the blocks "Stonehenge".⁽⁸³⁾ Another evident proof of the existence of intermediary walls, in addition to the remains of the walls already mentioned, are the chisel cuts engraved on the lateral faces of the pillars and still perfectly visible. In these incisions there were formerly inserted or connected the blocks of the intermediary walls.

The principal object of these cracks was to make the stones of smaller size which made up the wall form a single piece with the pilasters; in this way was supplied the necessary support so that in case of a sudden movement of the subsoil, the intermediary walls would not get out of level. In that period no mortar or other adhesive substance was used to join the component pieces of the construction as is the case in the modern technique and thus those mortice holes were indispensable. In addition to this system, as we shall see later on, they used metal bolts (Fig. 14) and stone wedges to fasten the smaller blocks together.

The majority of the pilasters which have remained on foot --- these are the "Kalasasayas" of the Aymara ---show on their tops, notches which indicate that there was a superstructure, or a construction the completion of which was planned. This presumption is confirmed by the suggestive fact that all the pilasters have varying heights (Cf. Pl. III), a circumstance which cannot be attributed only to the effect of their different bases, since of course the pilasters were certainly not built on a main foundation extending the whole length of the wall, as would be the case today.

As has been pointed out in previous chapters, one of the most interesting details of the balcony wall, is the distance between the center of each of the pilasters. In this, as has been said before, there has been revealed the normal measuring unit of Tihuanacu, or the unit of longitude, serving the purpose of the meter among us:

the "LOKA" of the Second Period, equivalent to 1 m. 63 cm.

This measure was used by the governing class, especially by the priests, who certainly were of the KHOLLA race. ⁽⁸⁴⁾ It is a question then of an ANTHROPOMETROLOGICAL measure.

Since this construction constitutes one of the most interesting details of the Kalasasaya of the Third Period and although a preliminary plan has already been published in the first volume (Pl. XVI) we believe it necessary to insert in the present volume a modern plan with greater details, for the better comprehension of certain points not treated in the former volume. The interior sloping of this wall to which we refer above, demonstrates a rather important point of construction.

This is that the wall made up of the pilasters was not vertically on plumb toward the upper part, but that it had --- and still continues to have --- a slant which at the present time is about two degrees toward the inside. This is equivalent to a diminution toward the top of some 38 mm. per meter, since within they are straight, or, let us say, vertical. In this way they possibly thought to have given greater stability to the wall of the construction. Therefore, the pillars are not parallelogrammatic, but become smaller toward their tops.

Although in the chapter dealing with the science of Tihuanacu and its probable age, we shall again discuss this notable wall, it is necessary in the present chapter to enter into greater details with regard to its construction and component parts still in existence.

There is not the slightest doubt that this wall was no ordinary wall and that its massive pillars did not serve simply to support it. Upon first observing this row of pillars, one might conclude that they were arranged without any regard for symmetry, since the distances from pillar to pillar vary. However, when one observes with greater attention, it turns out that the builders were not concerned at all with the space from pillar to pillar, since these were not of the same width, but the distance between the center of one pillar and the next was surprisingly the same in all cases, allowance being made, of course, for the slight dislocations of the wall attributable to geotectonic factors, which made them change position, though in a very insignificant manner, as can be seen in the following table. Thus the original space from the center of one pillar to another, allowing for some error of ours, would be, taking an average of the figures from the table, 4 m. 84 cm. 5.75 mm.

Neither is it impossible that these differences existing in the distances between the center of one pillar and another were intentional, for reasons as yet not clear to us, since they may have had some connection with the sunsets, considering that the six months between the spring and fall equinoxes have a different cycle than that from the fall to the spring equinox.

Table: Pillars of the Balcony Wall

The above very suggestive table, shows that the ten distances hover around four meters eighty centimeters. The average of the distances from one pillar to another, we repeat, or 4 m. 84 cm. 5.75 mm. would have corresponded to three arm spans, or three Lokas of the Tihuanacu of the Third Period, or a Loka of 161 cm. 3.25 mm., making a total of thirty Lokas from the center of pillar "A" to the center of Pillar "K". ⁽⁸⁵⁾

If one compares the foregoing measurements with those given in volume one, it will be noted that the Loka shows a difference of 1 cm. 6.75 mm. The cause for this difference of about a centimeter and a half per Loka is that to be expected between measurements not completely exact, made in that period (more than thirty years ago) with imperfect instruments, and the new and careful measurements carried out with a steel compensated for temperatures. ⁽⁸⁶⁾ To avoid errors, the readings are made in the following manner: The center of each pillar is marked; the metal tape is stretched from where pillar "A" begins to the end of pillar "K" and in this way the following result is obtained, which we judge to be unquestionable: total length 49 m. 30 cm., with the following figures from the center of one stone to the other.

The distance from the center of pilaster "A" to that of "K" is 48 m. 45 cm. 7.5 mm. which, divided by thirty normal measurements would give the figure of 161.51 which would be the true average, or, in other words, the "meter of Tihuanacu" of the Third Period. In my opinion, the distances between the center of each pillar to that of the next were regular during the period of the construction of the great wall, and the differences of a few centimeters which they show now come from movements in the earth, of which we have spoken before.

However, as we have pointed out, there may exist the remote possibility that the differences we have indicated came about intentionally for astronomic reasons, a fact which, in our opinion, is not completely improbable. It may be that the center of pillar "F" is not exactly the center of the balcony wall, since the center of the latter is 24m. 65 cm. and the distance from the center of pillar "A" to the center of pillar "F" is 24 m. 41 cm. 5 mm. and consequently the distance from the center of pillar "K" to the center of pillar "F" is 24 m. 4 cm. Herewith we present another table which shows the setting of the sun each thirty days toward the pillars "but" taking the center of pillars "A" and "K" as solstitial points, or as the maximum oscillation of the sun in its amplitude.

Since the balcony is so important by virtue of being the most monumental and significant part of the temple, we are presenting forthwith the dimensions in width and thickness of each pillar which forms this enormous south wall.

Since the wall of which these "kalasasayas" formed a part ⁽⁸⁷⁾ was not concluded, as the notches indicate, it is to be supposed that on the top of each pilaster, there was a superstructure with a little window similar to the one still supported by the north east pillar of the "sanctissimum" (Fig. 25), through the center of which more or less the sun set every month.

As was pointed out in another part of this volume, it seems that the main level of the interior of Kalasasaya was that of the platform of the perron, and there possibly correspond to this level those projections or ledges shown by some blocks or columns on the interior side of the building (Vol. I, Fig. 4), especially the southeast pilaster, which constitutes the corner of the east and south walls.

As can be noted in the different outlines of the general map of the temple (Map III) , the walls are not aligned now as they must have been in their time, nor are the pilasters on the exact level that they had when the building was in construction or in use, because, owing to the lack of foundations which we have mentioned, the majority of them have settled to a greater or lesser degree.

It seems that the building that we have been calling the "sanctissimum", must have had wholly or in part, a higher level than the interior, a fact evident from the greater height of the "observation block", which in turn represents at the present time the highest point of the interior in question, as can be seen clearly in the side view B. B. of the general map of Kalasasaya.

Something very suggestive in the interior which we are describing, is that more or less at the level of the superior platform of the monumental perron, or rather, at the point of altitude which in all the palace corresponds to the level of this platform, there is found a thin layer of white clay-like sediment, a sort of kaolin or fresh-water lime.

This is a fact which suggests two possibilities : either that this white layer is a sediment from the lake, formed when the waters which invaded the site of the temple dried up, or what is more probable, the material of which this white layer is composed, was carried there by the hand of man to form in the interior a snow white floor, well stamped down, and one which would have given the inside of the building an extremely attractive appearance. Perhaps it was planned as the base for a floor in mosaic work similar to the one found upon excavating the perron (See Fig. 23) --- a floor to be laid when the building was completely finished.

Various remnants of constructions which escaped vandalic destruction, indicate that in ancient times there existed different levels or platforms in the interior of the temple. This appears to be so especially in the section which extends from the "observation block" toward the balcony wall in a westerly direction and precisely in the region of the line where one finds the excavation furrow opened by the Crequi de Montfort Mission in 1903 and more or less in the center of Kalasasaya.

As we have pointed out before, there was found in this furrow, in addition to the carved longitudinal blocks of considerable size, a monolithic stair of two steps. This, without the least doubt, belonged to an intermediary construction between a lower and higher level, or possibly, at the approach to the platform toward the "observation point" of the Third Period, the spot destined for the erection of the Sun Door.

But the most notable thing found a few feet more to the west of the aforementioned steps and on the very intermediary line of Kalasasaya, is the beautiful block which in the first volume we called "PEDESTAL" (Vol. I, Pl. XV c). This, in our opinion and on the basis of its configuration, must have served as a foundation for the rustic observation instrument which in another section we called a "SIGHT." Its position in use was of course the opposite of that seen in photograph No. c, which is shown clearly by the opening for a sort of "drawer" inside, and by the typical staircase decoration on its upper part. ⁽⁸⁸⁾

Some seven meters farther on and in the same westerly direction, also almost on the perpendicular of the building and approximately on the line of the center of the west wall of the Second Period, there was found a gigantic slab, called by the people the "TABLE" and a short distance away was found the "Pedestal" (See Fig. 20, the position of both at the time of excavation). Its purpose was, in our opinion, (because of its smooth and slightly convex surface), to serve as a polishing stone, and for this use it was dragged over the surfaces of the columns of the balcony wall while these were still lying on the ground waiting to receive the final handiwork. The finishing consisted in a polishing with water and sand, a mixture introduced between the block and the polishing slab.

In considering the Third Period of Kalasasaya, it would also be necessary to refer to the repairs carried out in this period on the constructions of the Second. As we saw before, the men of the more developed, later periods, took advantage of the works of their predecessors by repairing and improving them. This happened not only with the idols, on which they placed ideographs connected with a symbology for a new and evolved theogony, but they also repaired the architectural works and adopted them to their needs and customs. One can observe this structural readjustment very clearly on the east wall of the Kalasasaya of the Second Period.

In this wall one sees a single technique and a uniform state of erosion throughout; but we also find inlayings, or, "repairs" with classic blocks and in a very superior technique. These blocks SHOW ONLY THE RELATIVELY INSIGNIFICANT WEARING AWAY OF THE MATERIALS OF THE THIRD PERIOD. Fig. 13a shows us, on the east wall of the temple, and in a very suggestive and clear manner, one of these many repairs with blocks wrought in a superior technique, besides the technique of the inferior period.

One of these well carved stones of the Third Period can be seen on the left side of the large pilaster which adorns the left side of the monumental perron. One can see there very clearly the enormous wearing away or erosion of this pillar and, one might say, the freshness of the block to its left. We believe it unnecessary to emphasize that the heavy erosion is not to be attributed to a material which was softer or more resistant to erosion. 

There also belongs to the Third Period of Kalasasaya the magnificent floor which, like a mosaic floor of a modern bath room, extends in front of the stair, as can still be seen in the photograph of the monumental perron taken at the beginning of 1904, immediately after this beautiful pillar was discovered by the French Mission. (Fig. 23). On the night following the excavation, a piece of this paved work was carried off and a few days later the whole piece was taken by the half-breed boys.

Thus the only testimony as to its existence as well as to that of the aforementioned "PEDESTAL" and a large number of other constructions of indescribable artistic value destroyed by them, are the magnificent photographs taken by the author and preserved almost intact in the archives of the Institute with which he is identified, as documentation for future studies. ⁽⁸⁹⁾ Possibly some of the red sandstone pillars of the north and south walls of Kalasasaya which were very much deteriorated, were replaced in the Third Period to preserve uniformity. This can be seen in the relatively insignificant erosion of some pillars which, we repeat, cannot be attributed to a more consistent material.

As the temple which is the object of our study was not completed in the Third or Last Period of Tihuanacu, neither were there replaced in this period the portions of the west wall which communicate with the balcony wall and which were built in the previous period. This wall (Fig. 13b), still extant, is done with such a crude technique that it might be compared with that of the First Primitive Period.
Although we are considering here chiefly the monumental Temple of the Sun, Kalasasaya, and although we shall consider it farther on, it is necessary to note the following with regard to the balcony wall of the Third Period.

Observing the setting of the sun from more or less 50cm. beyond the west edge of the Block, which we have decided to call the observation block, and looking toward the pillars of the balcony wall, the priest-astronomers observed that the sun set between pillars "K" and "A" about the solstices; as for pillars "B" and "J" they observed that the sun set between them about a month before and after these times. The same thing occured with pillars "I," "H," "G," "F," "E" and "D" more or less every month. We need not emphasize more the importance that this information had for the determination of agricultural seasons and connected celebrations. (Cf., in this respect, the diagram of Fig. 2, the angles of which can also be applied to the pillars of the balcony wall.)

What we have just noted is of fundamental importance for the following chapter, which deals with the astronomical angles of Kalasasaya and the age of this magnificent building of American man, unique in the whole world.

Many more pages could be filled in describing and studying the remains of this construction; but those who come after him who initiated these investigations, will study with greater care and competence all that concerning the mysteries hidden in this most noble work --- the most important, in our opinion, of all those which have remained in Tihuanacu and perhaps in all America.

(89) Institute "Tihuanacu" de Antropología, Etnografía y Prehistoria, Pinilla 556, La Paz, Bolivia.