Cold War Brazilian Armor

CCL X1A2 Carcará Brazil (1977-1978)
Light Tank – 24 Built

At the beginning of the 1970s, the Brazilian Army started developing armored vehicles. They would start with wheeled vehicles. After having successfully developed the prototype concepts which would become the EE-9 Cascavel and the EE-11 Urutu, the Brazilians looked to tracked vehicles. Like the previous wheeled vehicle projects, the engineers started small. They first set off remotorizing readily available M3 Stuarts, and then started developing the vehicle that is known as the X1 light tank. The X1 was a modernization of the Stuart which was armed with a low-pressure 90 mm gun and would be developed into an entire family of vehicles.

An attempt to improve the X1 by fixing some of its design flaws was unsuccessful. The X1A1 was developed to improve on the X1, but in the process only got worse. It was too long and too narrow, which made steering a very difficult task. An extensive rebuilding program would have been required to bring the X1A1 to a usable state, something which was simply not worth it. Considering that both the X1 and X1A1 used the now 30 years old M3 Stuart as their basis, some of the flaws would never have been able to be fixed because of the age of the vehicles.

As a result, it was decided that the development of a completely new tank was the way forward. Capitalising on the experiences gained from the X1 and X1A1 projects, the designated X-15 project would use components and design principles from the previous conversions. It would, for example, use the suspension of the X1A1, but also the X1A1’s turret for the first prototype. The resulting tank of the X-15 project would be known as the X1A2 and be the first (and so far only) serially produced tank which was fully designed in Brazil and used in active service.

The X1A2.
Source: Image Caiafa

The X15

It is suggested that the development of a new nationally designed light tank began quite early. The exact date is unknown, but it might very well already have started with the development of the X1 in 1973, and might have really started to take steps after the failure of the X1A1. Somewhere during the project’s life, the vehicle received the X1A2 designation, most likely when it was decided that the X15 would use components from the X1 series.

The X1A1 in 1976.

Nevertheless, the Centro de Pesquisa e Desenvolvimento de Blindados (CPDB) (English: Centre for the Research and Development of Tanks), and the Instituto de Pesquisas e Desenvolvimento (IPD) (English: Research and Development Institute) were studying a light tank concept with two main goals in mind. The first was to abandon the overhaul and the conversion process of the M3 Stuart, which was a laborious task and reached its limit for the goals of the CPDB. The second goal was to create a vehicle which was able to steer properly.

The resulting project was designated X15, with the 15 referring to its planned weight of 15 tonnes (16.5 US tons). To save costs and time, the engineers decided that it would be best to take advantage of the efforts already made by integrating components from the X1 projects in the X15 design. The suspension and turret of the X1A1 were carried over, alongside other components. It supposedly had the same Scania diesel engine as the X1 and X1A1, but with improved horsepower from 260 hp to 280 hp.

The hull would keep many features from the X1A1, but feature an improved armor design with better ballistic shapes for the front hull. The X1A2’s hull was also wider than its X1 predecessors, from 2.4 meters to 2.6 meters (7.9 to 8.5 feet). This widening of the hull would result in a significant improvement in the overall mobility of the X1A2. The X1A2 also featured an Allison CD-500 transmission instead of an M3 Stuart or 18-ton M4 tractor transmission.

CD-500 transmission.
Source: TM 9-1730B – Maintenance Cross-Drive Transmission model CD-500

The X1A2 prototype

The exact date for when the first X1A2 prototype was finished is unknown, but there is proof that the X1A2 prototype hull was nearing completion in July 1978. Considering that a Deputy Chief of the Brazilian Army suggested the interruption of the X1 Pioneiro production for the X1A2 in July 1978, it can be reasoned that the X1A2 prototype was built between July 1977 and July 1978. As mentioned, this prototype integrated the suspension, turret, gun, and engine of the X1A1, while also using new components and design features to fix the issues of the X1A1. It was tested by the Parque Regional de Motomecanização da 2a Região Militar, (PqRMM/2) (English: Regional Motomecanization Park of the 2nd Military Region), after completion. After testing, the vehicle seems to have been accepted and the design of the production version was initiated.

The X1A2 prototype with the French DEFA D-921 gun. Note IPD and CPDB written on the side.
Source: Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives

The production version differed in some significant ways from the prototype. While the prototype seems to have had a 4-man crew, considering 2 sets of sights were installed on both the driver’s and co-driver’s side, the production version only had a set of sights for the driver. In addition, the hull machine gun was also removed. The now vacant space of the co-driver was supposedly filled with additional 90 mm ammo racks. Besides the removal of the co-driver role, the production version X1A2 was also armed with an EC-90 gun.

The EC-90 was a license-produced low-pressure 90 mm gun by Engesa. This gun was based on the Cockerill Mk.3 gun. The switch from the French DEFA (Direction des Études et Fabrications d’Armament) (English: Directorate of Armament Studies and Production) D-921 gun to the EC-90 had multiple reasons. The main reason was that the French company SOFMA (Société Française de Matériel d’Ármament) (English: French Society of Armament Materiel) would only sell their D-921 guns together with the turret, while the X1 family used a local design. The license production of the EC-90 gun made the X1A2 cheaper to produce. In addition to manufacturing costs, the Brazilian Army only operated their EE-9 Cascavels armed with the EC-90. The adoption of the EC-90 on the X1A2 would simplify logistics as well.

The same X1A2 prototype from a different angle, note the two sets of periscopes.
Source: Blindados no Brasil


The X1A2 was produced in two production batches, with the first consisting of 10 vehicles and the second of 14 vehicles. Of these batches, only the first would enter active service, while the second batch mostly ended up as gate guardians and monuments. The first batch X1A2 was officially designated as Viatura Blindada de Combate – Carro de Combate MB-2 (VBC CC Medio Bernardini-2) (English: Armored Fighting Vehicle – Combat Car Medium Bernardini-2), while the second batch was designated as Viatura Blindada de Combate – Carro de Combate MB-2A (VBC CC Medio Bernardini-2A) (English: Armored Fighting Vehicle – Combat Car Medium Bernardini-2A). The reason for this difference was because the second batch used more locally produced components and featured a swing arm for the .50 machine gun, instead of a fixed mount. Interestingly, this swing arm seems to have been incorporated in the X1A2 prototype, but not on the first production batch.

Considering the X1’s similar designations, it would most likely have also been referred to as the Carro de Combate Leve X1A1 Carcará (CCL X1A1 Carcará), (English: Light Combat Car X1 Carcará), but this is more of an educated guess that cannot be actually confirmed. The Carcará was an indigenous crested hawk and was previously the nickname of the X1A1. The nickname most likely carried over from the X1A1 to the X1A2 because the X1A1 project was unsuccessful, and the X1A2 carried over many aspects from the X1A1.

A Carcará hawk.

The first production batch was delivered to the 6th RCB in Alegrete, Rio Grande do Sul State, where they would replace a squadron of M4 Shermans. The 6th RCB was the only unit to ever operate the X1A2.

The swing arm.
Source: Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives


For the construction of the X1A2, multiple parties and companies were involved. The most important company which built the X1A2 was Bernardini. Bernardini initially manufactured truck bodies and value transport vehicles, and came in contact with the Brazilian Armed Forces by manufacturing trucks for the Brazilian Marine Corps and the Army. With Bernardini being a manufacturer of safes and armored doors, they were requested by the Brazilian Army to help build the X1. After the X1 was successfully developed, Bernardini started developing the X1A2 together with the PqRMM/2 and the CPDB engineers.

Company/Army Component(s)
Bernardini Most likely: hull, turret, engine installation, equipment installation, track mounting and suspension
CSN Steel
Novatração Tracks
DF Vasconcelos Periscopes
Scania-Vabis Engine
Engesa EC-90 90 mm gun
PqRMM/2 Design support and testing
X1A2 during ramp tests at the PqRMM/2.
Source: Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives

The X1A2

Multiple characteristics are presented incorrectly in the book Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives and the Tecnologia Militar Brasileira Magazine. As a result, all the length values were calculated with ratios and are reasonable estimates. Most of the other values are rough estimates. The X1A2 weighed 19 tonnes (21 US tons) and the hull was about 6.06 meters (19.8 feet) long, 2.6 meters (8.5 feet) wide, and 2.45 meters (8 feet) tall. It had a crew of three, with the driver located on the front left of the hull, the commander/loader on the left side of the turret, and the gunner on the right side of the turret.

An X1A2 of the 6th RCB on a 7 September parade in Alegrete.
Source: Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives


The hull of the X1A2 was a completely new design with an improved ballistic shape on the front hull, compared to the Stuart based X1s. The overall hull design still bore much resemblance to the M3 Stuart in general design aspects. The thickness of most of the armor is unknown but is expected to not differ too much from the original M3 Stuart thicknesses. The frontal armor is said to be 50 mm (2 inch) thick but it is unknown if this is for the entire frontal hull. Most of the upper hull was protected by a 50 mm plate angled at about 60 degrees from vertical. On top of that was a small strip of armor right behind the drives periscopes, which slightly extended upwards. The thickness of this plate is unknown, but might also have been 50 mm angled at about 60 degrees from vertical. The lower hull plate is also estimated to have been 50 mm thick and was angled at about 45 degrees from vertical. Under the lower hull plate was another lower hull plate which was probably the same 50 mm thick but angled at about 60 degrees from vertical.

The side, rear, top, and floor armor are unknown, but probably did not differ much from those of the original Stuart, and are thus estimated to have been 25 mm thick. The rear armor is also estimated at 25 mm thick (1 inch). The top plate is estimated to have been about 13 mm thick (0.5 inch), and the floor plate between 13 mm to 10 mm thick (0.5 to 0.4 inch). What is known about the bottom plate, is that it is a dual plate against mines. These values are nothing more than the most likely estimations.

The X1A2 had a headlight on both fenders and a horn on the left fender. It seems that the large upper hull plate was also the access hatch to the Allison CD-500 transmission. It could be bolted loose and subsequently lifted from its position. The sides of the hull were used to mount pioneer tools and were also composed of integrated storage boxes. The X1A2 presents a large lifting hook on both sides of the side engine plate at the rear. On the left fender was presumably another storage box and the exhaust was located on the right rear fender. On the top rear hull were two hatches to access the engine and, in front of those, what looks like an air inlet grill for the engine.

The front hull, note the hatch for the CD-500 transmission.


The X1A2 was powered by a Scania-Vabis DS-11 6-cylinder in-line 280 hp diesel engine. It used a three speed Allison CD-500 transmission, the only vehicle of the X1 family to use a different transmission than either the M3 Stuart or 18-ton M4 Tractor transmissions. The X1A2 had a top speed of potentially 60 km/h (37 mph) on-roads, 30 km/h (18 mph) off-road, 15 km/h reverse (10 mph), and an on-road operational range of 600 kilometers (373 miles) and off-road of 350 km (217 miles).

The X1A2 used a copied and altered Vertical Volute Spring Suspension (VVSS) system of the 18-ton M4 artillery tractor. It had 6 road wheels divided over three bogies, with 3 bogies per track, 3 return rollers on each side, a drive sprocket in the front and an M4 Sherman idler wheel on the rear. The newly designed 18-ton M4 Tractor/M4 Sherman hybrid suspension gave the X1A2 a ground pressure of 0.63 kg/cm2 (9 psi). The vehicle could climb a 0.8 meter (2.6 foot) vertical obstacle, and a hill at an angle of 40 degrees. The X1A2 had an on ground track length of about 3.66 meters (12 foot) and could cross a trench of 2.1 meters (6.9 foot).

The 18-ton M4 Tractor/M4 Sherman hybrid suspension. Note the M4 Sherman idler and the three sets of bogies.


The X1A2 turrets were practically the exact same turrets as the X1A1 turret. The front turret was armored with 25 mm (1 inch) thick steel plates at various angles to protect it from .50 caliber machine gun fire at 200 meters (218 yards). The rear parts of the turret were armored with 12.5 mm (0.5 inch) thick steel plates. It is suggested that the overall turret layout and the internal turret construction and components were more or less copied from the French H-90 turret. It had the exact same turret ring and its overall shape seems to match the H-90. In addition, in the first BT-90 turret of the X1, a lot of equipment was carried over from the H-90, like the periscopes.

The X1A2 turret had a fixed mount for a .50 cal machine gun on the left side of the turret, in front of the commander’s cupola (the second batch had a swing arm mount). The commander’s cupola’s structure was slightly raised from the turret top to provide the commander with a 360 degree view. The antennas of the radio sets were located behind the gunner’s cupola on both sides of the turret. Spare tracks were mounted on the turret bustle sides, which might hint that the turret bustle sides were 12.5 mm (0.5 inch) thick, as it required additional tracks as armor. This placement of the spare tracks meant that the smoke dischargers were moved to the front of the turret, in a set of 3 dischargers on each side. A small light was also installed on the turret side of the commander’s cupola. On the very rear of the turret was a storage box welded on the turret rear, right behind the spare track protected plates.

The turret could traverse 360 degrees in 12 seconds with a hydraulic drive. The base rangefinding is done through graticule rangefinding, but laser and coincidence rangefinders were offered as well.

Side of the X1A2’s turret.


The production X1A2s were armed with the license produced EC-90 90 mm low-pressure guns manufactured by Engesa. These guns were derived from the Cockerill Mk.3 guns. The low-pressure gun allowed vehicles like the X1A2, but also the 5 tonnes AML-90, to mount a gun with significant armor penetration capabilities.

The EC-90 gun schematics.
Source: Manual de Opercão 9110-733-604 – Torre ET-90 II e Armamento

The trade-off was that these guns would, for a long time, only be able to fire High Explosive Anti Tank (HEAT) ammunition because Armor Piercing (AP) rounds simply had too little muzzle velocity to compete with HEAT. Armor Piercing Fin Stabilized Discarding Sabot (APFSDS) rounds did later appear, but these did not provide any improvement in penetration compared to HEAT rounds. A 90 mm APFSDS round for the later Cockerill guns would penetrate 100 mm (3.9 inch) of armor at 60 degrees from vertical at a range of 1,000 meters (1,090 yards), compared to 130 mm (5.1 inch) at 60 degrees for HEAT at any range. This meant that the APFSDS round mainly served as a round against targets with HEAT countermeasures.

Engesa did develop an APFSDS round for their EC-90 gun on the EE-9 Cascavel on request from Iraq around 1985, but this round would never finish development and only a few test batches were ever produced. The main issue was that the APFSDS round needed to reach higher velocities than the low-pressure rifled guns were designed for. The fixation which kept the sabot together would get damaged when the round was fired. Another issue was the muzzle brake, as the then-current muzzle brakes of the EC-90 guns prevented the use of APFSDS rounds. A pepper pot style muzzle brake was used to solve this issue, but the downside was that the pepper pot muzzle brake was less effective in mitigating recoil. Why the development of the APFSDS round was eventually cancelled is unknown. An Engesa engineer who worked on the APFSDS round believed that they would have been able to fix the problems considering the round started development 8 years before Engesa’s bankruptcy. The employee also stated that the project just did not take off and interest in the round from Iraq was probably not large enough to complete its development.

An Engesa employee with an APFSDS round of the EC-90 at Engesa’s test range.
Source: Private collection

Contrary to common statements, the X1A2 was thus not able to fire APFSDS rounds in the configuration it used at the time. Not only were they not available at the time when the X1A2 was in service, the development of the APFSDS round was never completed by Engesa. It also did not have a muzzle brake that would support the APFSDS round. Since neither requirements were met, the X1A2 never used APFSDS in its loadout. In addition, by the time the APFSDS round could have been ready, interest had already completely shifted to the M41C and the main battle tank projects like the Osorio and Tamoyo.

The X1A2 had access to HEAT, High Explosive Squash Head (HESH), and High Explosive (HE) rounds. The HEAT round was meant for anti-armor purposes and was the X1’s anti-tank round. The HESH round was mainly meant for bunkers, walls and light vehicles, and not as ‘anti-armor’ ammunition. The HE round was used as a general purpose support round. The X1A2 also had access to a white phosphorus smoke round and a HEAT practice round.



Effective range


HEAT (High Explosive Anti Tank)

130 mm (5.1 inch) at 60 degrees from vertical or 350 mm (13.8 inch) flat at any range.

2,000 meters (2,185 yards)

900 m/s

HESH (High Explosive Squash Head)

Meant for bunkers, walls and light vehicles.

2,000 meters (2,185 yards)

800 m/s

HE (High Explosive)

Lethal radius of 15 meters (16 yards)

2,000 meters (2,185 yards)

700 m/s

HEAT-TP (High Explosive Anti Tank – Training Projectile)

Training projectile

2,000 meters (2,185 yards)

900 m/s

White Phosphorus – Smoke

Smoke round

2,000 meters (2,185 yards)

690 m/s

The X1A2 stowed 24 rounds in the turret and an additional 44 rounds in the hull, for a total of 68 rounds of 90 mm ammunition. In addition to the 90 mm, the X1A2 mounted a turret top .50 caliber machine gun (750 rounds) for the commander, and a coaxial .30 machine gun (2,500 rounds). It has a gun depression of 8 degrees and elevation of 17 degrees. The X1A2 had 16 smoke grenades for its 6 smoke dischargers.


The X1A2 was delivered to the 6th RCB in January 1981, with 10 X1A2s replacing the M4 Shermans of the 2nd Tank Squadron. The X1A2s operated together with the X1s of the 1st tank squadron, to which the X1’s were delivered in 1978.

X1A2’s of the 6th RCB.
Source: Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives

The fact that the X1A2 replaced the M4 Sherman and had a larger turret than the X1 led to a very interesting situation. Being used to the 3-man turret of the M4 Sherman, the fresh X1A2 crews tried to adopt the same practice in the X1A2 turret. The turret was very cramped and the practice was abandoned. According to veterans, the commander would have to exit the turret and re-enter the turret to use the radio in the turret bustle. In a real battle situation this would have been impractical and dangerous.

X1A2 crosses a bridge laid by the XLP-10 bridgelayer.
Source: Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives

The X1A2 would encounter various issues during its service life, with the 18-ton M4 tractor torque converter being the biggest issue. The torque converter used by the tank was not designed for a vehicle of the size and speed of the X1A2. What made matters worse was that it was lubricated by poor quality oil used in Brazil. The high concentration of sulphur and low flash point caused the component to wear out much quicker.

To fix this issue, Bernardini suggested replacing the 18-ton M4 torque converter with a TwinDisc converter from the US. Bernardini would acquire one torque converter from TwinDisc and it would be successfully tested. Bernardini ordered enough torque converters to refit the X1A2 fleet, but due to the M41C program, they were never installed.

The controlled differential also caused issues for the X1A2. The more wear the differential had, the harder the X1A2 became to steer. An additional downside compared to the M41 Walker Bulldog was that the entire turrets of the X1 family had to be lifted from the hulls to perform maintenance to the drive shaft of the tanks. Another issue that caused premature wear and difficulty in operation was the lack of instruction manuals for the X1 family as a whole.

X1A2 of the 6th RCB and three X1’s.
Source: Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives

Export Attempt

In the early 1980s, the Brazilian government and Bernardini attempted to export the X1A2 to Paraguay, which at the time only operated 21 M3 Stuarts and 3 Sherman Fireflies (the Stuarts being gifted to them by Brazil (12) and the United States (9), and the Shermans by Argentina, eventually replaced by 3 Sherman Repotenciados). As a sign of good will and in an attempt to make the Paraguayans more favourable towards the X1A2, the Brazilian government offered the revitalization of 15 M3 Stuarts. The M3 Stuarts would be upgraded to the X1P standard by receiving a general maintenance overhaul and the Scania-Vabis engine. These Stuarts are still in active service in the Paraguayan Army. Eventually, the X1A2 was never bought, potentially because the Paraguayans wanted to acquire the EE-9 Cascavel instead, of which 28 were delivered in 1985 together with 12 EE-11 Urutus. The X1A2 had a unit price of around 400,000 US Dollars in 1980 (about 1.3 million US dollars in 2021), against 243,600 US Dollars for the EE-9 in 1988 (about 560,000 US Dollars in 2021).

The X1A2 in Paraguay.
Source: Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives


The issues the X1A2 had, in addition to the upcoming M41C upgrades, would cause the Army to refrain from deploying the second batch of X1A2s. They were stored in São Paulo, where the vehicles continued to deteriorate until they were eventually discharged from service in 1989. A couple of these vehicles were turned into gate guardians and monuments, but others were scrapped.

The X1A2s would be gradually replaced from 1988 onwards by the M41C. The X1s, and probably X1A2s as well, were decommissioned in July 1994.

The M41C.


The X1A2 was an interesting step for the Brazilian defence industry. It was the first and so far only serially produced, albeit only in limited capacity, tank fully designed in Brazil to see service in the Brazilian Army. It had its issues, but most of these seem to be fixable or were almost fixed by Bernardini. The only real issue the X1A2 would have is the differential, but since it had an improved length to width ratio compared to the X1A1, the steering was already much better. The X1A2 was a promising vehicle if these issues were fixed, and more importantly, if it was not overshadowed by the M41C program.

If the X1A2 was developed a bit earlier, it would have most likely seen more service, and its early flaws would have been fixed. Considering the start of the X1 family only began in 1973 and the X1A2 was only developed from 1976 onwards, while the M41 upgrade programs started their development in 1978, it seems that the first successful attempt of developing a national tank was inevitably too late. The X1A2 is the logical end to the development of Stuart based light tanks with 90 mm guns, which started in 1973. The Brazilians tried to design their own tank and succeeded. From the experience of the X1 program, Bernardini started the development of the M41 upgrade programs and the development of Brazil’s first Main Battle Tank: the MB-3 Tamoyo.


The X1A2 Prototype. Done by Brian Gaydos.
The X1A2 production vehicle. Done by Brian Gaydos.
The X1A2 production vehicle with camouflage. Done by Brian Gaydos.

Specifications CCL X1

Dimensions (L-W-H) 6.06 meters (19.8 feet) long including the gun x 2.4 meters (8.5 feet) x 2.45 meters (8 feet) tall
Total weight 19 tonnes (21 US tons)
Crew 3 (Driver, Commander-Loader, Gunner)
Propulsion Scania-Vabis DS-11 6-cylinder in-line 280 hp diesel engine
Suspension Bogie suspension
Speed (road) 60 kmh (37 mph)
Operational range 600 kilometers (373 miles)
Armament 90 mm D-921 low-pressure gun
.50 machine gun
.30 coaxial machine gun


Front (Upper Glacis) 50 mm (2 inch) at 60 degrees
Front (Lower Glacis) 50 mm (2 inch) at 45 degrees
Sides (guess) 25 mm (1 inch)
Rear (guess) 25 mm (1 inch)
Top (guess) 13 mm (0.5 inch)
Floor (guess) 13 to 10 mm (0.5 to 0.4 inch)


25 mm (1 inch) frontal part
12.5 mm (0.5 inch) rear part

Production 24
Special thanks to Expedito Carlos Stephani Bastos, the leading expert in Brazilian vehicles, please visit his website for further reading on Brazilian vehicles:, Jose Antonio Valls, an Ex-Engesa employee and expert in Engesa vehicles, Paulo Bastos, another leading expert of Brazilian Armored vehicles and the author of the book on Brazilian Stuarts and the website, Adriano Santiago Garcia, a Captain in the Brazilian Army and ex-company commander on the Leopard 1 and ex-lecturer on the Brazilian Armored School, and Guilherme Travassus Silva, a Brazilian with whom I was able to endlessly discuss Brazilian Vehicles and who was always willing to listen to my near endless ability to talk about them.

Brazilian Stuart – M3, M3A1, X1, X1A2 and their Derivatives – Hélio Higuchi, Paulo Roberto Bastos Jr., Reginaldo Bacchi
Blindados no Brasil – Expedito Carlos Stephani Bastos
Jane’s Light Tanks and Armoured Cars of 1984
Worldwide Tank Fire-Control Systems – CIA
Personal correspondence with Expedito Carlos Stephani Bastos
Personal correspondence with Paulo Roberto Bastos Jr.
Caiafa Master
Engesa brochures and manuals
Cockerill brochures
TM 9-785 18-Ton High Speed Tractors M4, M4A1, M4C, and M4A1C – US Army April 1952.
Stuart: A history of the American Light Tank, Volume 1 – R.P. Hunnicutt
Tecnologia Militar Brasileira magazine

Leave a Reply

Your email address will not be published. Required fields are marked *