The seals of the seven angels are quite extraordinary in character and shape, quite like the rest of the manuscript itself.
Seal of Michael (Sun)
The first seal is meant as a general illustration at the beginning of a new book in the manuscript, as the title announces: Incipit Incipit Liber Tertius Clavicula Salomonis (Here Begins the Third Book of the Clavicula Salomonis). It s flanked by the astrological siglae of Mercury on the right and Moon on the left, but we know it well to be the seal of Michael from the Heptameron.
Studying both seals, we see that none of its characteristics are accidental, so the direct source is not a printed Heptameron. From left to right: the beginning of the horizontal line, in the large curve, is not a circle or a closed loop, but a mere line or a slightly descending curve; Where the H seals had a descending slanted line, both these seals have a curious broken like. This is a quite new characteristic in these seals. In the U shaped part of the seal, where the H seals have a cross usually, there is nothing here. This is consistent with one source: MC.3, the Wroclaw Magical Calendar. The rest of the seal is quite normal, but we are keen to examine the loop: in the H seals, this loop is not a minor trait, its a rhomboidal shape. No other version has a round shape (and MC.3 has nothing at all) except for MC.1.
Seal of Gabriel (Moon)
The seal of Gabriel also bares great resemblance to the other seals, but it has quite a few unique characteristics.
The first vertical segment is reduced to a simple line, disjointed from the main horizontal line. The second vertical segment has very simplified heads, in the shape of rigns or semi-rings. One of the most intriguing features is the lack of horizontal compartments in the enclosed figure: only one vertical line, unlike any seal encountered. While we would expect it to be similar to the Scot seals and the MC series, it has exactly what they lack (a vertical line) and it lacks exactly what they supply (the three horizontal lines). The seal maintains the H feature of the two lower case b-s, as they lack in the MC versions altogether and are not morphed into 6s. The following element, the A shape, is much more consistent with the MC series, as it is cursive, unlike Scot and the H series. Another prominent feature would be the presence of two cursive Ms separated from the main line, not one angular M jointed to the line, a feature completely new. The last swirl is quite consistent with the H series, but much more cursive, resembling more a cursive U. If the seal of Michael would lead us to a manuscript akin to CM.3, the Wroclaw Codex, but not identical, the seal of Gabriel also seems to support that.
The seal of Samael
As accustomed to surprises, this seal does not fail to disappoint. From the minor morphology changes in the previous seals, this one takes a leap. The first element is a smallish simple cross, like all other versions. The next element is no longer a straight angle with a loop, but a loop upon a horizontal angle, the whole element is slanted to the right. The next section of the seal is completely changed, retaining only small resembling elements: the x shape and the horizontal line through the circle remind us of the previous seal variations, but the circle is not at a 45 degree angle, its placed horizontally, from it protruding two symmetric elements with angles below and small arrows joined at the top. This seal has a completely new look and cannot even be counted as a copyists error. It is unlike anything I have met with in the study of these seals.
Seal of Raphael (Mercury)
The source of this seal in its oldest form has to be the Heptameron, as the attribution of the seal is not interchanged with that of Michael. Yet again, novelty and change. The first element is a backwards L, but not started from the medial line, like most seals. The second element is similar, a circle breaking the line, with a vertical line through it, but it is simple and unadorned with a foot, like most have it. The following common element, a simple line through the medial one, is missing. The third element of this seal, where most figures would have an elongated circle with a cross inscribed, turn here into a sort of beetle-shape: a vertical ellipse halved, crowned above and below with two semicircles, a unique trait. The right half is equally original: we have one X shape instead of two verticals and two crescents, on one side and the other of the X. The final K shape is fairly standard.
Seal of Sachiel (Jupiter)
The seal of Sachiel is entirely different from the others and seems to be independent. Not only the morphology of the seal is entirely alien to the other Heptameron-based seals, but the seal itself is unknown. The top part of the seal is a horizontal line, capped with tho descending oblique crosses, the middle being joined to the top of an angle. The ascending angle is capped with two rings and it contains a symbol similar to a 2, possibly derived from the astrological sigla of the planet Jupiter. The lower half is some sort of graphic construct meant to describe a flaming sphere s bottom, perhaps the result of a vision obtained in the practice of the rituals prescribed. The flaming sphere bottom has a trapezoid attached, azured, with flames or flame-like symbols as well. To my knowledge, this is completely original, and has not been encountered anywhere.
Seal of Anael (Venus)
This seal is entirely original as well, and not previously attested anywhere. It has no resemblance with the Heptameron seal. This character is made up of a trapezoid seated upon an elongated ellipse, flaming beneath. It is capped with two rings in the upper corners and halved by a horizontal line, itself capped with seals. The upper part contains what seem to be two incomplete Venus siglae, joined to the top line.
Seal of Cassiel (Saturn)
Conclusions: the CSS is a highly ecclectic mix of seals, some taken from traditional sources, some modified as the writer saw fit, and others entirely original. If the author used the Heptameron, he did not consult a printed version, or chose alternate seals from another copy of it or a similar work, undiscovered.