The Villa and Tomb of Lucullus at Tusculum

Author(s): George McCracken

Source: American Journal of Archaeology , Jul. - Sep., 1942, Vol. 46, No. 3 (Jul. - Sep.,
1942), pp. 325-340
Published by: Archaeological Institute of America

Stable URL: https://www.jstor.org/stable/499772

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/499772?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and

extend access to American Journal of Archaeology

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 THE VILLA AND TOMB OF LUCULLUS AT TUSCULUM

PRESERVATION of the names of at least thirty-six owners of Tusculan

republican period is eloquent testimony of the popularity of that region
wealthier Romans.' While under the empire the number of known pr
only twenty-nine, this drop is probably to be explained more by the smalle
of surviving literature of the period than by any real decline in the numbe
or by any substantial shift of favor from the ager Tusculanus to other reg
three writers of the Silver Age express preference for other sections, but t
haps no more than a personal taste,2 since presumably most of the 13
villa sites described by me 3 were in use in imperial as well as republi
Of them all, none was ever more famous than Cicero's Tusculanum, the
which, indeed, subsequently passed into literature as a synonym for a p
country estate. Yet, in spite of the many and ingenious attempts made t
through five centuries of antiquarian interest, its exact site still defies iden
and we must await further discoveries before we can abandon our agnos
on this point. Moreover, thus far we have been able to associate with
site the name of no republican owner-all owners heretofore definitely
with their villas being of imperial date.
After Cicero's villa, to judge only from frequency of reference in the lite
Tusculanlumrn of the conqueror of Mithridates occupied proximos honores,
on account of its relative magnificence, it did not really surpass Cicero
size and grandeur. To attempt to identify the villa of L. Licinius Lucullu
fore the purpose of this article.
Plutarch (Luc. 39) tells us that at Tusculum or near it Lucullus had a
(the plural apparently does not mean more than one villa) with belved
corridors, and cryptoporticuses. As most of the larger and many of the sm
must have had at least some of these features, there is no assistance he
mining the site. It was at this villa that, on a visit to the conqueror, Pompe
him because he had well prepared his house for summer, but had not equipp
winter. Then Lucullus replied with a laugh, "Do you think, then, that
intelligence than the storks and cranes, that I do not change my residen
seasons? "6 It would seem clear that the villa was at least, in part, open
side air, and more so than the average villa, or Pompey would not hav
remark. We should expect as much from Varro's statement (RR. 3, 4,
villa was so designed that it was possible for Lucullus to dine on some b
1 George McCracken, A History of Ancient Tusculum, Washington, 1989, pp. 368-386, he
as History.
2 Stat. Silv. 1, 3, 84-85; Mart. 10, 30, 5 f.; Plin..Epp. 2, 17, Q, but Hor. Epod. 1, 29 f. is not a case in
point, since Horace is merely comparing a Tusculanum with Maecenas' favor.
3 Iistory, pp. 206-307.
4 History, pp. 377-385; G. McCracken, "Cicero's Tusculan Villa," CJ. 30, 1934-35, pp. 261-277.
As lead pipes seem not to have been in use as early as Cicero's day, I no longer hope to hear of one found
with his name on it.
G Gelzer, RE. s.v. Licinius Lucullus no. 104 (25, cols. 676-414). I Varro RR. 3, 17, 6-9.

THE ARCHAEOLOGICAL INSTITUTE

OF AMERICA

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 8~6 GEORGE McCRACKEN

triclinium and to see others flying i

based on personal observation, for V
own,8 but was also the adoptive fath
possessed intimate knowledge of the
Casinum one much larger (ib. ?, 5, 8-
size of the villa and of its picture-galle
a villa. One would have expected rather
On the size of the villa we have also
sors criticized it because there was le
may have been on this occasion that
luxury.1" He said that a knight lived
their villas were also luxurious, he th
those beneath him in rank. The rem
Cicero, but even if the story is fictitio
of necessity be true. While Scaevola's
mella (1, 4, 6) tells us that Lucullus' w
the excellent library of the villa aft
To these data we should perhaps add
fish-ponds or piscinae to be seen on som
from which, when Cato became the gu
he sold enough fish to bring the sum
type are not found in large numbers at
reservoirs described in my book being
it was at the villa near Naples that th
ever, may have sold some from the T
reason to believe that this villa was p
Thus far, we have no topographical
the one fact about the excessive size
as a whole. No evidence of an aviary h
is the type of structure that would be
description of ancient aviaries given
tures were perishable. The other deta
We must now give our attention
'On this type of aviary see A. W. Van Bur
MAAR. 5, 1925, pp. 111 f.; 10, 1932, pp. 10-1
8 Cic. Fam. 9, 6, 4; 9, 2 and 5; Varro RR. 3,
M. Licinius Lucullus, afterwards M. Terent
no. 109 (25, 414-419). 10 Cic. Leg. 3, 30.
1 F. Grossi-Gondi, II Tusculano nell'Etd Classica, Roma, 1908, p. 101, wrongly states that this was
Gabinius. See T. Ashby, PBSR. 5, 1910, p. 249.
12 Cic. Balb. 56; Att. 4, 16, 3; De Or. 1, 24-27; History, p. 373, no. 18.
'~ Cic. Fin. 3, 2; Acad. Pr. 9, 148; Isidorus Or. 6, 5, 1.
14 History, pp. 387-416. 15 Ashby, PBSR. 5, 1910, p. 250.
16 The text is that of the Loeb editor, C. E. Bennett, edited by M. B. McElwain (1925), but revised
by me in the light of the photograph of Codex Cassinensis (C) printed by Clemens Herschel, The Two
Books on the Water Supply . . . of Frontinus2, New York, 1913. The date of C is variously given as
from the eleventh to the thirteenth centuries, the latest Teubner editor, F. Krohn (1929) declaring for
the twelfth. For the De Aquis this codex is all important, since our other MSS were all derived from it.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 THE VILLA AND TOMB OF LUCULLUS AT TUSCULUM 327

Concipitur 17 Appia in agro Luculano Via Praenestina inter miliarium septimu

sinistrorsus passuum septingentorum octoginta.
Cn.'s Servilius Caepio et L. Cassius Longinus, qui Ravilla appellatus est, censo
ditam sexcentesimo vicesimo septimo [197 B.C.], M. Plautio Hypsaeo M. Ful
aquam quae vocatur Tepula ex agro Luculano, quem quidam Tusculanum credun
adducendam curaverunt. Tepula concipitur Via Latina ad decimum miliarium
Roma dextrorsus milium passuum duum ....... inde suo rico in urbem per
Idem [= Agrippa]'1 cum tertio consulfuisset [927 B.c.], C. Sentio Q. Lucretio consul
tertium decimum quam luliam deduxerat, Virginem quoque in agro Lucullano c
... Concipitur Virgo Via Collatina ad miliarium octavum palustribus locis, sign
darum scaturiginum causa.

If the statements made in these sentences are right-I assume

correctness of the text-then we have here a remarkable coincidence. Three of the
nine Roman aqueducts are thus represented as having their sources on the estate
of Lucullus (so our translations), and that in spite of the fact that their respective
sources are located by Frontinus on three different Roman highways. At the time of
the building of the first two aqueducts (Appia 31U, Tepula 1U5 B.C.), our Lucullus,
the most famous of those who bore the cognomen,20 had not yet been born, whatever
date we accept for his birth, and before the third was built in 19 B.c. he had been
dead for at least thirty-five years. When Frontinus wrote his treatise about 100
A.D., Lucullus had been dead for a century and a half, during much of which time
we have no evidence for the existence of any of this name. The persistence of the
name, even of a Lucullus, in connection with a private estate through such a long
period would be remarkable indeed, considering the number of owners who must
have possessed it in the interim, but to find three such properties, each retaining the
name of Lucullus so long after his death, and each the source of an aqueduct, is ex-
traordinary, to say the least.
In the first passage quoted, the intake of the Appia is located on the Via Prae-
nestina between the seventh and eighth milestone on a cross-road to the left or north,
780 paces from the highway, while the third places the Virgo's intake on the Via
Collatina near the eighth milestone. A glance at any map of the course of these
roads 21 will reveal the Viae Collatina and Praenestina at this point running roughly
parallel, not an English mile apart. Since the Praenestina is to the south, a point
780 paces north of it would be close to the Collatina also, in fact probably even
nearer the latter. It is therefore abundantly clear that the intakes of these two
aqueducts actually are not far apart on different roads but really quite near each
other, though Frontinus chose to give their locations by referring to different roads.

17 Frontinus De Aquis 1, 5: Lucullano edd.; luculano C.

18s Ibid. 1, 8: Lucullano edd.; luculano C; dextrorsus C, but dextrosus Herschel wrongly; Krohn has
dextro+sus in another hand, but the r is quite clear, and the second hand is doubtful. After dextrorsus
a blank long enough for four letters, and another before inde. The latter is shown by Herschel and
Bennett as a lacuna, probably because they thought Frontinus had given the exact number of feet. I
believe, however, that the blanks are due only to improper spacing of letters added to a blank space
either by a corrector or the original scribe. The word rivo is an addition of the editors where C has a
blank space of one letter's length. 19 Ibid., 1, 10: lucullano C.
20 Ten Licinii Luculli are known. See RE. 25, cols. 372-419, nos. 99-110 (no. 105 is fictitious). All of
them are republican. Curiously enough, Lucullus' son is not listed among them.
21 See Gori's map reproduced in part by Herschel, opp. p. 184.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 328 GEORGE McCRACKEN

Perhaps on one of his inspection tr

and on another the Virgo by the Vi
erence to his routes, forgetting for
In the nineteenth century the sourc
springs which he found north of th
first passage cited, to read Collatin
the most expert, but also the most re
of the springs proposed by Lanciani
fore Lanciani must be wrong in his id
volume has subsequently changed,
that to accept Lanciani's emendatio
Frontinus mistaken.
Thus, it is quite possible that the intakes of both Appia and Virgo were on the
same estate and this was called ager Lucullanus, whatever that may mean. We ordi
narily find the word ager used with place names -I have found no examples with a
personal name, but I see no reason why it could not be so used. I should, however,
prefer to translate it as "an estate of Lucullus" rather than "the estate of Lucullus,"
as we know of other villas of Lucullus elsewhere.24 No other evidence for this particu-
lar estate is known,25 but it can not have been the Tusculanum on account of the
distance. It might be the villa for winter use implied by Lucullus' statement about
the storks and cranes, but I should prefer to consider that villa the one near Naples
This brings us to the source of the Tepula, which has been identified with the
modern Sorgente Preziosa, a spring rising about one and a third Roman miles due
west of Grottaferrata, just north of the large bend of the Marino railway line.26
The spring, now in use, is covered with a building and can not be inspected, but it
has been noticed by others that the temperature of the water is considerably higher
than that of the surrounding air. When the air was only 470 Fahrenheit, the water
was 61-630, an indication of the reason for the name. Another confirmation is the
fact that the spring lies at about the proper distance - more exactly if the lacuna b
recognized in the text (supra, note 18) -from the tenth milestone of the Via Latina,
approximately at the modern Villa Senni which is built on the site of the Vicus
Angusculanus or Res Publica Decimiensium, the first statio of the Via Latina.27
Moreover, there was an ancient deverticulumn which left the Via Latina between the
tenth and eleventh milestones (rather nearer the latter) and went in the general di-
rection of the Sorgente Preziosa.28 Because this deverticulum left the Latina near the
eleventh milestone (Borghetto), Rocchi wishes to emend the numeral decimum to
undecimum.29 Confusion might easily arise between X and XI, but if Frontinus gav
the extra feet beyond the even miles, then the tenth milestone would be a better
reading.

22 R. Lanciani, "I Comentarii di Frontino," AttAccadLincei ser. 3, 4, 1880, p. 247.

23 T. Ashby, The Aqueducts of Ancient Rome, Oxford, 1935, p. 51.
24 Cic. Off. 1, 39, 140; Varro RR. 3, 17, 6-9, and the passages cited above.
2, Ashby, Aqueducts 49, note 1. Nothing on Lucullus' villa in connection with the source of the
Virgo (167). 26 History, fig. 47 and Grossi-Gondi, folding map of ager Tusculanus.
27 History, p. 196. 28 History, fig. 47 and pp. 176-177.

29 A. Rocchi, "I1 Diverticolo Frontiniano," Studi di Storia e Diritto 17, 1896, p. 15.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 THE VILLA AND TOMB OF LUCULLUS AT TUSCULUM 329

Whichever numeral be correct, the identification is certain and c

be given to Lucas Holste (or Holstenius), who visited the re
1649.30 The following passage is quoted by Ashby 31 from on
scripts (Cod. Dresd. F. 193, f. 43r+v): 32
Inspexi fontem Tepulam, vulgo nunc la pretiosa dicta; est in valle Marciana sub
in Via Latina ad XII lapidem; in dicta valle ad Crabram est oticina ferraria, u
passus scaturit fons aquae copiosissimus, vulgo la Pretiosa dictus, quam Tepul
tissimum est, distat enim duobus m. pass. a decimo (vulgo le Murene) dextrorsum
Frontinus neget Tepulam certum habere fontem sed ex venis collectam, existim
rivatas, postquam Juliae dictu receptae amplius in urbem fluere desierunt, hoc
Quod etiam idem Frontinus Tepulam agro Lucullano concipi ait, id huic font
villae Luculli maxima extant vestigia sub Burgetto ad sinistram viae Latinae, ub
per vineas aliquot porrectas inspexi: ab hisce vestigiis villae Lucullanae DCC circ
ille Pretiosa dictus; puto tamen multo longius se protendisse agrum Luculla
pontem Crabrae sub Decimo ubi immensa illius villae vestigia visuntur quae vulg

If the Sorgente Preziosa be really the source of the Tepula, an

tification correct, then the spring would have been, according
property of Lucullus' Tusculanum. With this view, however, there
ties. The site is in a region not particularly suitable for an esta
cence, since the elevation is relatively low, the highest point
being only N00 meters above sea level, compared with the 384
Aldobrandini at Frascati, to choose a typical Renaissance Tuscu
ple.33 Even now there are few modern residences near the spring,
of a half mile I have been able to discover definite traces of but a
and, at just that distance, doubtful traces of another, neither o
site at all imposing.34 I find it hard to believe that a villa as m
Lucullus could have existed in this area, later to disappear so co
not in modern times be even traced. Quite large structures, to
peared in Rome since the Renaissance, but here in the open Cam
ings have not been erected and it is difficult to see for what p
wilfully destroy bulky antiquities, except to gain material for
We must examine Holste's identification of the villa. He say
Burgetto ad sinistram viae Latinae" and that the remains he s
paces from the spring. I take it that he means on the left as o
for both he and Frontinus locate the spring with this orientat
his details are inconsistent. If the ruins were on the left, or n
Latina, they must have been more than a full Roman mile from
there are a number of villa sites within his 700 feet, all of the

30 The same conclusion was reached independently, it would seem, by Fat

Lanciani, Comentarii 294: G. Tomassetti, La Via Latina nel Medio Evo, Rome
31 PBSR. 5, 1910, p. 222, repeated in English in Aqueducts, pp. 159-161.
32 The Castello Savelli, or Borghetto, lies athwart the Via Latina almost at
Le Murene here represents Villa Senni, not the modern Casal Morena at the nin
is really villa 35, a magnificent establishment, but one that had ample land abo
33 Elevations from the Carta d'Italia of the Istituto Geografico Militare, fol.
34 Indicated on my map (History, fig. 47) but unnumbered, as I believe the
culanus.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
330 GEORGE McCRACKEN

i.e. downhill from the castello, the

also of its modern counterpart, the
the Latina.35 The remains which H
since their ruins would satisfy all
Holste probably made no exact meas
mentions no positive evidence, epi
Lucullus. His view rests only on th
a faulty relationship at best. Moreo
ment of a large villa on a small pro
There is still another difficulty, nam
no trace of a monumental tomb, fo
Lucullus in the Campus Martius, hi
interment to be made on the Tusc
might have disappeared, for it would
tombs which have been seen in rece
a trace, but reasons which will appear
sis until all possibilities are exhaust
These difficulties have led me to s
scrutiny, and to suggest two interpre
difficulties. The first presupposes t
the site of the Sorgente Preziosa can
this side of the ager Tusculanus. Inde
I limited my work on Tusculum to a
Marciana, a stream which runs nort
the gorge below the Badia di Grott
north of the spring. If this line be co
lanus, but the whole of the valley
be saying that the Tepula was brou
of some, was in the ager Tusculanus,
other testimonium concerning this v
site was Tusculan. As this is unsatis
must have lost much of their earlie
passage, particularly since it does no
The other interpretation assumes t
I am satisfied that the codex Cassinen
agro Luculano, although the initial
takably lu. At least they are not go
where in the same script, but resem
fessor J. B. Titchener, the abbreviati
meant to be vz as if for lu, and I t
since I find ex agro videlicet Culano d

35 Villas 90, 91, 105-107, all relatively unim

3 Plut. Luc. 43. B. Gatze, Ein rimisches Ru
also to Mart. 7, 3, 5, an error for 8, 3, 5, an
7 History, pp. 156-160.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 THE VILLA AND TOMB OF LUCULLUS AT TUSCULUM 331

made by the scribe of C of the words ex agro Tusculano, sin

lanum credunt immediately follows. What I believe Frontin
aquam quae vocatur Tepula ex agro Tusculano Romam et in
curaverunt. Influenced by in agro Luculano or Lucullano - it
the preceding passage, a scribe who produced one of the an
illegible Tusculano and wrote Luculano instead. This would
t to 1 and the omission of s. The scribe who did this could not have been the one who
made C, for the relative clause would then be unexplained. That was added later as
a marginal gloss by some one who had a copy of the De Aquis reading ex agro Lucu-
lano Romain etc. Such a person may have been one of Frontinus' successors as prae-
fectus aquarum, but he need not have been better informed about the aqueducts than
any intelligent reader of the statements immediately following and in the next
chapter. There it is abundantly clear that the tenth milestone of the Via Latina is
certainly near the ager Tusculanus-in it, in fact, though Frontinus does not ex-
plicitly say so. In a later copy, perhaps in C itself, the scribe inserted the marginal
gloss as an interpolation.
There are two advantages in this reading of the text: (1) the coincidence is re-
moved, and (2) the difficulties of finding the villa of Lucullus near the spring are
avoided; while as a by-product the entire Valle Marciana is shown to be in agro
Tusculano, but the statement of Frontinus really gives us no help in locating the
villa. Fortunately, we are not without still another clue that may prove fruitful in
our search, and that is the approach from the point of view of the tomb.
Within the ager Tusculanus there have survived to our own day two tombs, both
associated with Lucullus. The first of these, known locally as " Sepolcro di Lucullo,''38
is at Frascati, on the piazza of the same name, a modest square on the southeast
edge of the city, at the head of the street which leads from the Piazza del Duomo
to the right of the cathedral facade. The tomb stood on the south side of the im-
portant dererticulum which passed this point on its way to Tusculum from the tenth
milestone of the Via Latina,39 the ancient road here coinciding with the modern
path to Villa Ruffinella. The remains are incorporated into a modern house, out of
which bulge the picturesque masses of concrete forming the tomb's core. Mattei 40
states that in 1598 (misprint for 1698, as the cathedral was built about 1700) the
tomb was despoiled of its decorations for the building of a new cathedral, but that
the better pieces were taken to Rome "da diversi Cavallieri Romani . . . per ador-
narne le loro Gallerie." Much the same story is told by Giovanni Antonio Pallotta,41
who also gives the detail that the pieces taken to Rome were of very fine marble,
among them statues.
Mattei gives a rough woodcut of the tomb as it looked in his time, and this is
copied by a better draughtsman in Vulpius' account.42 Angelini and Fea publish a

3s History, p. 431, no. 24. 39 Ibid., p. 165 f.

40 D. B. Mattei, Memorie Istoriche dell'Antico Tuscolo oggi Frascati, Rome, 1711, p. 62, reprinted in a
Latin version by S. Havercamp in Burmann's edition of J. G. Graevius' Thesaur. Ant. & Histor. Ital.,
Leyden, 1723.
41 Quoted by Fra Domenico of Frascati in a MS. cited by Lanciani, BCAR. 12, 1884, p. 211.
42 J. R. Vulpius, Vetus Latium Profanum, Rome, 1742, 8, 80, tab. 2, no. 5.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 33t GEORGE McCRACKEN

fairly accurate drawing of the build

showing the exterior square with pier
cre concrete, with aggregate of selc
Latin cross, faced with reticulate of s
terior is based on Ashby's accounts,
that the veneer is described as mar
in the interior, makes it seem probab
case, we have no positive evidence th
Nibby 45 must reject the identificatio
travagances of local patriotism. It
close to a very important Tusculan
which underlies so much of Frascat
be sure, all of the evidence for dati
known proprietors being successive
mitian, but the site must also have be
hand, the tomb is on the opposite side
hardly satisfy also the requirement of
there was extensive ground, as wit
the destruction of Tusculum in 1191 A.D.
Similarly, the name of Lucullus has been attributed to the great tomb known
locally, from the family that has owned it for upwards of sixty years, as "Torrione
di Micara." The structure was built on the north side of the same important deverti-
culumi on which the other tomb was built,47 at precisely the spot where an alternative
route diverges to the northeast to join the first farther on at Capo Croce outside
Frascati. The modern lane passes the tomb on the north, but the ancient road ran
to the south, so the tomb stood in the intersection of the two roads, and doubtless
made an imposing appearance to travellers coming from the direction of Rome.
In mediaeval times the building was made into a fortress, as were so many ancient
tombs, Guelphic battlements being added. Eschinardi 48 calls the structure the
"Isileo nella vigna Rocci" and says that for a time it served as a novitiate for the
Badia di Grottaferrata. A plan was published by Uggeri 49 and repeated by Angelini
and Fea 5o and by Canina.51 Mattei (108) prints a rough figure which is probably
meant to represent this tomb, as he mentions the building on the next page. In it
the battlements are only crudely indicated, if at all, and on top of them something
appears which may be meant for thatching, but this is also far from clear. The bat-

43 G. Angelini and A. Fea, I Monumenti pii2 Insigni del Lazio, etc., Rome, 1898, 2: Via Latina, tav. 7.
Most copies of this rare work lack the second volume.
44 L. Canina, Descrizione dell' Antico Tusculo, Rome, 1841, p. 135, tav. 26 =History, fig. 138, and view,
tav. 7 = Canina, Edifizi Antichi dei Contorni di Roma, Rome, 1856, 6, tav. 83 = History, fig. 145.
45 A. Nibby, Analisi Storico-Topografico-Antiquaria della Carta de' Dintorni di Roma2, Rome, 1842,
3, p. 597, cited as Analisi.
46 History, pp. 250-260. Villa 84 will be discussed later. 47 Ibid., p. 169.
48 F. Eschinardi, Esposizione della Carta Topografica Cingolana dell'Agro Romano, Rome, 1696, p. 369,
also edited by R. Venuti, Descrizione di Roma e dell'Agro Romano, Rome, 1750, p. V75.
49 A. Uggeri, Journdes Pittoresques des Edifices de Rome Ancienne, Rome, 184: Journke de Tusculum,
tav. 5. 50 Angelini and Fea, !2, tav. 4.
"~ Canina, Descrizione tax. 26 = Edifizi Antichi 6, tav. 82 = History, fig. 138.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 THE VILLA AND TOMB OF LUCULLUS AT TUSCULUM 333

tlements could, of course, not have been added after Mattei's t

graph taken from the south, with a telephotographic lens,52 is
while Grossi-Gondi publishes three views of the exterior and o
The construction is of cut peperino, badly weathered, the blocks
high, rusticated and with false joints. Nibby's measurements a
m., base to plinth 2.37 m., plinth 0.72 m., plinth to cornice 4.74
m.54 The inside diameter is 26.68 m. and the interior is empty exc
and three chambers on the southwest side, now enclosed in a m
which has been locked whenever I have visited the site. These
Ashby,55 of brickwork with a corridor in front of them, and in t
way leading to the terrace above. The chambers follow the cu
but the northeast wall is straighter. The foundation of the tom
to the southwest where it is rectangular. This was on the sid
traveller from Rome would approach, but, curiously enough,
there, but on the north, and not, it would seem, oriented in relati
way. When the building came to be used as a fortress, the pos
doubtless caused a more practical relocation of the roads, so tha
directly to the door. On the southwest side of the exterior is a pla
which I believe was placed there in antiquity and may have bor
tion. It is too high for measurement, but it is the same height as
and about 2.50 m. long, i.e. approximately the same size as the
is set. It faces the ancient highway and an inscription on it c
from the road. A similar block of marble was inserted into the wa
Caecilia Metella on the Via Appia.56 Other parallels between
striking. Her tomb, built of travertine and sperone (lapis Gabinus)
bers of brick, is dated by Frank not earlier than 10 B.C., on the ba
of bricks, but he recognizes a chronological difficulty in this
Metella,67 the wife of M. Licinius Crassus, would have been at l
age at this time. This makes him think that the erection of t
taken some time after her death. If, however, he is wrong ab
brickwork, then her tomb may be earlier.
On this point we get help from Goitze's recent study 58 of the c
architectural form. He limits the vogue of this type to the pe
the middle of the first century B.c. to some time before the end o
the earliest examples being the tomb of Caecilia Metella and the
Others of the group include the tombs of Munatius Plancus at
at Ponte Lucano on the Via Tiburtina, of the Lollii TTUrbici at Con
52 PBSR. 4, 1907, pl. 14, fig. 1= History, fig. 139.
63 Grossi-Gondi, tav. 1 (see fig. 1), fig. on title, heading of p. iii, tav. 5 (see f
R. Lanciani, Wanderings in the Roman Campagna, London, 1909, p. 293; G.
setti, Tusculana, Rome, 1912, p. 7. Figs. 3-4 show the exterior.
4 Nibby, Analisi 3, 356; also his Viaggio Antiquario in alcune citta del Lazio:
i.e. MS. notes in the possession of Ashby.
55 PBSR. 4, 1907, p. 134.
56 CIL. 6, 1274; T. Frank, Roman Buildings of the Republic, Rome, 1994, p. 14
57 See RE. 25, col. 268 f., no. 56 (Miinzer); RE. 3, col. 1235, no. 136.
-8 B. Gatze, Ein r6misches Rundgrab in Falerii; Baugeschichte des rimischen A
Grabbau und Bauplanung des Augustus, Stuttgart, 1939, reviewed in CW. 34

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 FIG. 1. -TORRIONE DI MICARA FROM THE SOUTHWEST (AFTER GROSSI-GONDI)

FIG. . --TORRIONE DI MICARA, INTERIOR (AFTER GROSSI-GONDI)

FIG. 3. -TORRIONE DI MICARA, DETAIL

OF EXTERIOR (AUTHOR'S PHOTOGRAPH)

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 THE VILLA AND TOMB OF LUCULLUS AT TUSCULUM 335

and unidentified examples at Casal Rotondo on the Via Appia

Vicovaro and the ninth kilometer-stone of the Via Labicana, a
lia and two at Bieda (15).
The tomb of Caecilia Metella he would date (10, 15) near th
tury, much earlier than Frank. Before the end of the centur
the use of which was restricted to the highest nobility, was
because of higher building costs, and also the end of rivalry b
ilies of the period. Though G*tze has apparently made no thoroug
dence for the identification of this tomb as that of Lucullus - he
hesitancy - on stylistic grounds the torrione fits extremely w
for the tomb of Lucullus, a building in which an attempt was ma
mausoleum in the Campus Martius (19). Lucullus died about 56
may well have been built some years later. It certainly could
during the lifetime of the conqueror, or the people of Rome
advance that he was to be buried at his Tusculanum. Since thi
this type anywhere in the ager Tusculanus, I feel certain that we
confidence that the traditional attribution is correct, and I c
my former skepticism on this point."9
Having established the identity of Lucullus' tomb, we may p
this fact will help us to locate the site of the villa as well; but fi
examine the various sites that have been suggested by others
As we have seen, Holste, on the occasion of his visit in 1649,
the villa, identifying it, as I understand his words, with the r
68 or 70, but this view rests only on the belief, probably mistak
Preziosa was on the Lucullan property. In addition to the ob
(p. 330), we may now safely say that these villas are too far from
serious consideration. The actual distance by the ancient road
Roman mile, and there are also two deep ravines and the sit
tween.

No evidence exists for the view which I find adopted by John Breval, an En
traveller of two centuries ago, that the Villa Aldobrandini is the site of Luc
Tusculactum, unless Breval derives his identification from the first tomb mention
above, which stood rather near that villa.6o Breval assumes that it was from t
villa that the Aqua Virgo got its water, citing Pliny 30, 3, which has nothing
on the subject. He is clearly referring to the statement of Frontinus, already
cussed, but wrongly attributes it to the Tusculanum of Lucullus. Breval can h
be referring confusedly to the view of Kircher, the only modern writer wh
cites, that the Lucullan villa occupied the site of the Renaissance Villa Torlo
since I know of no reason to believe that the Aldobrandini or Pamfili familie
owned it.

The Villa Torlonia, also called Caravilla,61 Galli, Borghese, Altemps, Ludovisi,

H9 History, p. 428. See also the skepticism expressed ibid., 405 f. on reservoir 65 (below p. 339).
60 J. Breval, Remarks on Several Parts of Europe, etc., London, 1738, 1. p. 81, note n.
61 R. Lanciani, Storia degli Scavi, Rome, 1908, 3, pp. 50-53; PBSR. 5, 1910, p. 249; History, pp.
300-304. This villa was not part of villa 59, as some have thought.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
336 GEORGE McCRACKEN

Poli-Conti, and Sforza-Cesarini, lies on

while little ancient material is now to
gether with much that has been seen
large and important villa in both repu
was first made by Kircher,62 and a
Lanciani,63 Grossi-Gondi (113-141), Be
view rests on Kircher's statement that "
Ludovisiorum est, uti ex inscriptionibus
verbis: L. LUCUL. LUC. F." (73). Is h
ments or of more than one copy? Lan
pipes for stones, and this is possib
rejects this inscription as spurious, an
Kircher says it was, one would expec
a type more appropriate to lead or br
date. If authentic, the provenance be
testimonium worthy of respect that
provenance, confirmation appears to be
owner, written on September 14, 156
"nel loco proprio di Lucullo che cost m
menti, e di alcune lettere che vi ho trova
Bertano on April fifth of the same y
Grossi-Gondi maintain, refer to the d
more than a century earlier.
Grossi-Gondi calls attention to tw
cinius P. 1. Philonicus and P. Licinius
seen by Fabretti "in villa nobilium d
in eodem loco in villa nunc Muti parieti
villa 81, stands southwest of this site
have disappeared. They may, as Ashby
Lucullus, but we really have no evide
where they have been seen (see below
freedman of a P. Licinius really refle
is another matter.
This identification rests, then, upon what is at best the confused testimony of
Kircher, supported in part by the statements of Caro, but the translator of the
Aeneid was admittedly interested in the prestige of his property and may not have
known the truth. It is, therefore, better to accept the judgment of the master of
Tusculan epigraphy, Dessau, that the inscription is spurious, particularly since the

62 A. Kircher, Latium id est, nova & parallela Latii tum veteris tum novi descriptio, Amsterdam, 1671,
p. 73. 63 Lanciani, Comentarii, p. 500, note 580; Storia degli Scavi 3, pp. 50-53.
64 U. Benigni, Cath. Encycl. s.v. Frascati.
66 P'BSR. 5, 1910, pp. 247-q50 and 302; see also History, p. 304, note 1. J. H. Westphal, Die riitnische
Kampagne, Berlin, 1899, p. 33, places the villa at Grottaferrata and E. C. Knight, the reputed author of
the anonymous work, A Description of Latium, London, 1805, pp. 142-150, locates it between Monte
Porzio and Grottaferrata, a very wide area. 66 Lanciani, Comentarii, p. 500, note 580.
*7 Lettere inedite di A. Caro, con note di Mazzuchelli, Milano, 1830, 3, p. 117.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 THE VILLA AND TOMB OF LUCULLUS AT TUSCULUM 337

site is too far from the tomb in any case. The distance is not
half Roman miles, with a very deep ravine and the sites of
81-82, 131) intervening.
Much nearer the tomb, however, indeed within two-thirds
stood three villas, two on the southeast of the ancient highw
and one on the northeast. The last is no. 131 at Casale Bevila
markable polygonal retaining wall on the northeast, but not m
there. On the basis of the relative unimportance of this villa, we
The same is true also of villa 75 which stood in a place call
down on the slope of the hill on the opposite side of the same
of a mile from the tomb. The platform, of good size, is now
mains, the results of the excavations of Pietro Santovetti in
unknown.69 Here, again, the comparative unimportance of the
ject it with confidence.
This leaves to be considered only villa 76 and in this case we
the remains of a magnificent villa of considerable size and ric
is known locally as Fontana Piscaro (perhaps a reflection of th
tioned below?), and when I visited it in 1931, it belonged to
Carletti of Frascati, but it has also been called the Vigna del
cia, and Vigna Varese.70 Kircher (72-76) was the first to speak of
a very faulty plan, which hardly does more than grasp the ge
much can be said for it, and this is reproduced by Vulpius.71 Kir
views of what he calls " Grotte di Lucullo, o, ii Centrone," but th
than views of the Severan palace on the Palatine! Ashby 72 di
in connection with the villa of Centroni (no. 35), but the pro
to the plan of the villa now under discussion and also Kircher's v
b)e mentioned later, make me feel certain that Kircher is pass
resentation of this villa, with the additional slip of inconsistentl
as "Centrone."
In Kircher's time the vigna belonged to the Varese, or Roccia family, and w
part of the property of the Villa Muti, but later fell to the Seminary at Frasc
Canina made a fresh plan of the remains which he gave to Uggeri to publish (ta
and this was repeated by Angelini and Fea (tav. 5), while later Canina publish
himself.73 Its great faults are (1) that it completely fails to show several of the
structions which must have been clearly visible in Canina's time, and (2) it sh
arbitrary reconstructions of the type characteristic of all Canina's work, as A
suspected. By penetrating into rooms filled almost completely with earth at p

68 IPBSR. 5, 1910, 245, pl. 25 =History fig. 125; Nibby, Analisi, 3.354; Viaggio Antiquario 1.7
plan 108; E. Dodwell, Pelasgic Remains in Greece and Italy, London, 1834, pl. 121; History 366 f
69 History, p. 289; PBSR. 4, 1907, p. 135 and addenda ad loc. in Ashby's possession.
70 History, pp. 289-293.
71 Vulpius 120, 128, tav. 4 = tav. 4 of Anon., Veteris Latii Antiqua Vestigia . . . praecipue Tyburt
Twusculana et Setina, Rome, 1751, reprinted by R. Venuti with slightly different title, Rome, 1776
72 PBSR. 4, 1907, p. 121.
73 Canina, Descrizione, p. 137. note 31, tav. .8 = History, fig. 78, but Fea labels the plan with his
name.

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
FIG. 4.--TORRIONE DI MICARA, DETAIL FIG. 5.--VILLA 76, LATER CRYPTOPORTICUS
OF EXTERIOR (AUTIIOR's PHOTOGRAPH) (AUTHOR'S PHOTOGRAPH)

FIG. 6. -VILLA 76, EXTERNAL RETAINING

WALL ON SOUTHWEST (AUTHOR'S PHOTOGRAPH)

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
 THE VILLA AND TOMB OF LUCULLUS AT TUSCULUM 339

where Canina's plan shows extensive details, I satisfied myself

picions were wholly justified.
The plan made by Ashby and Newton 14 I found accurate in e
as I was able to check it. Little of the upper part of the villa re
plot and by the time of my own investigations even that had
vegetation. The substructions, however, show this villa to have bee
est and most complicated in the region of Tusculum. The total
by 560 feet, reveal the immense size, while the actual substruc
the platform measure 585 by 220 feet. Two distinct periods of
distinguished, each possessing a cryptoporticus, the later being
and magnificent (see fig. 5. Fig. 6 shows the exterior faqade on
some of the underground rooms I noticed traces of wall painti
have here a villa which answers all of the requirements of th
literary testimonia. No aviary has been found, but this would hard
excavation. Villa 75 is, to be sure, nearer the tomb, but villa 76
requirement of a building too large for its site. The villa exten
to a point very close to the ancient road and on the opposite sid
siderable depth. It would have been impossible to extend the v
in either direction, while to the southeast the ground rises sh
Muti. The other remaining side faces the tomb, but between th
stands a circular open reservoir 45.70 meters in diameter. Ashby
six feet deep, but when I examined it, the depth was only half
ment of the diameter is accurate to the third decimal, and the con
circle, then when filled to a depth of six feet, the pool would h
792,000 gallons. It has often been published, first by Kircher w
"Sub villa Varesiana piscina sive Colymphydra Lucullani sub hac
and shows the outlet and stairway no longer visible, if they ev
probably referring to this piscina, when he says: "Part of the vast
Amphitheatre and Palace are yet to be seen between Frescati [s
place called Grotta di Lucullo." How Breval, who could certain
have misread Kircher's caption is difficult to understand, but t
him, the phrase Grotta di Lucullo, and the nature of Kircher's view
certain that Breval got his information from Kircher's book, wh
he appears to have visited the region. Because of the position
reference to the villa, I believe that it formed part of the sam
fish were sold from it by Cato.
The inscriptions of the two freedmen of P. Licinius, already men
either at Villa Muti, one time part of the same property as vi
nobilium de Rocciis" which may also be this site called Vigna
whatever force their discovery has for identifying the villa of Luc
to the site under discussion, rather than to Villa Muti.
If, indeed, the testimony of Kircher and Caro is, though gar
significant, the inscription may have been on lead pipe as Lan
7 PBSR. 4, 1907, pl. 15 = History, fig. 79.
15 Kircher, p. 74; Vulpius, p. 128. f.; tav. 5= Anon., Veteris Latii Antiqua

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms
340 GEORGE McCRACKEN

for Lucullus' villa would naturally h

Torlonia is considerably higher th
the former to the latter. On the
at Tusculum or elsewhere are imp
in the reign of Tiberius, so Kir
earliest example known, if authen
In any case, the nature of the rem
make me now certain that at last
crlanum with confidence, as well a
of the other, and, since no records
both sites would well repay excav
earliest Tusculanum, the owner of
rectness. To Kircher belongs the c
went astray in identifying the sit
views he definitely associated th
Lucullus.
GEORGE zMCCRACKEN
OTTERBEIN COLLEGE

"7 I have discussed the evidence for this statement in a paper: Miscellanea Tusculana, part I, read by
title at a meeting of the American Philological Association in 1941 (see TAPA.72).

This content downloaded from

5.179.1.3 on Tue, 31 May 2022 08:26:09 UTC
All use subject to https://about.jstor.org/terms